REPORT ONR-CR213-136-2F

'ULVE2,>1

I COLOR DISPLAY DESIGN GUIDE

ByM. J. Krebs

IJ.D. Wolf DDG* I H. Sandvig

CA. HoneywellC -Systems & Flesearch Center CI [2600 Ridgway Parkway NE

Minneapolis, Minnesota 55413

N00014-77-C-0349

OCTOBER 1978FINAL REPORT FOR PERIOD 15 APRIL 77 - 31 DECEMBER 78

Approved for public release; distribution uilimited.

PREPARED FOn THE"

OFF iCE Or NAVAL RFE-SEARCH*800 N. QUIN:.Y ST.0 ARLINGTON CVA 622.217U k'A'. I. '

GRICINAL COt.TAIN3 COLORl P1. Tvr.$: ALt DVWC• ILI) 3 (1)~ p REPRODUCTIONS VVttL BE IN Lii.ACK AND WHITE

CHANGE OF ADDRESS

Organizations receiving reports on the initial distribution listshould confirm correct address. This list is "located at the endof the report. Any change of address or distribution should beconveyed to the Office of Naval Research, Code 221, Arlington,VA 22217.

DISPOSITION

When this report is no longer needed, it may be transmitted to otherorganizations, Do not return it to the originator or the monitoringoffice.

DISCLAIMER

The findings in this report are not to be construed as an officialDepartment of Defense or Military Department position unless so"designated by other official documents.

,"I .1.1

z4

I'N( CI- .\ I 1:1:1) _ _______

ýLL kill I Y LA' % IFý FI IFN 0i I l4IS A (WkICN UAIA LNTLRLOF

REPORT DOCUMENTATION PAGE READ INSTRUCTIONSBEFORE COMPLETING FORMA

I H I It (Ii I N um ht $~442 OV -T A C C E ~S S IO N N U M UE H 4 I I E .IlI 't N I S C A TA L OG N U M ~l 4 4

4. TI~ILL (ANL.SýUtlIII t) 5. TYPI OF 4-dPOF4/PLI-4IO) LOVLPLI,

i ( Fimal R~eport, 15 Al1iZ l9')7-i

1 1%.1 Nrch'l 1.1. Sanrdvig I1,

1) Ptý HF OIMINk4 0HUANIZA11iONS NAW /AOORLS~-, I(0 PROGRl.AM t LEMENT PfmLI)C I A',9 A;L A

I 1J ~\**~ 111(iis HrdIescarJY' Cen tter &WORK UNi NUMB PS

2 iOO l~ dpxw;iY 'i~rkwayTN i [II'IP AI~j N 5541'1

1 H ON I 01LINZG OFF ICL NAML/401L)RESS -

Oiffice of Natval !l~eseairel, Code 212ý 0cwI"I

8100 N. Qirincv St rout al ___________I Arlirigion, VA 22217 TV2LSS:1C 0IN O

14.~~ -1 . SECOIYCLASSIFICATION (NOF 74 1(N QtC(H TI )1

16 IISTHIBUT ION SIIATCMENT (OF T141, IkFA k~.. ___ ______

\pprov ed tJJi1jgrelease: di s;trihtition unlirnit 12(.

!ILI1hI~IU1 19j t~ALE 6mjaQM.B ~ tO4f8O*~t. "FPR N F r ýiEPORT)

* -~ -- -ORIOINAL CONTAINS COLOR PLATES: AU. DW

-' F REPRODUCTIONS WWL BE IN 9A.ACK AND W1J)Wr

Mr~]NCPKY -E D ; VYD BBLOC K N U-M UC ())

C olor codin g M~ulticolor displays1I-Jt tronic displa1ys, liiformation displays

\ircrn¶ dis-pflays

AUSRAC (0N NUON NEVEF(SE SID IF-NECESSARY AND IOTNiTICfl'YI ~ Y~it oj( ct'iv lV( XV' to d1evclop adesign guide for the useu of color in advainced aivionicsy displays. Study results are presented in two p1rti-I. ] 1art I presentsL priniciple~s for Ius-C

* of color, plus supporting d~ita xhvre such dat a exist. P1rinciples of I airt I arc ¶penera1land can lie ipplije _nvrulyay applicaition involving color coding o;' t'is pl~ infnr -

<mat ion. I 'art 11 repre~sents, the application of these priniciples~ to real- -wurid k ockpit*displatys. 1Reconirnended color codes akre dev-ý-oped for aipplication to t-lectronic dis-

plays in t'ipht r/al~t ak aircraft. 1-xarnples of representative displayý formatsIDF124 L 47 REDITllION OF ( NOV 55 'S ')dSOL-E E LI(LSSILI _________

5ECURITY CLASKIFICA-TION OF THIS (AGL (Wil ý4N Al A tF iRNI (

UNC LASS IFIECD11 [mTV CLAr ;IF:ICATION OF THIS PAGE qWHEN OATA ENTERIEO

20. Abstract (continued)

incorporating the recommended coding are provided. K .

. -.. I I" C I_\

II

*1

I•N (.' ASS IF 11:1) oS I •U RIVI ' C L A Sr'S l •I [ \ I' }rU N U Fr 1i , '~Is ' .T . IVY N aI• , 'I N,[ • I - • T -

Th is color dpaydesign guide was written for use by a human factor.--

engni'r. Ioevtc ill at temipt has bet'n made to express theý. concepts

and p~r ni1' iples inl 13lng1rage understandable bY the genieral reatder initerested

inl tire effective applicaition of color to a displaNv

It is uissuiried that the reaider is vith ex considering the use of color inl a

pairticular dis-playN or h as been tasked with det ermiining how anl existing

color capabhility might lie used e'ffectivelY it, somei specified application. Asnf thiz Ori,vif~tcinn thlc rinvilmotit fi#C 'es ~ -M ý

jthe relative merits of color versus various other codes. Rather, it is IaSsu4111Id

thant a decision to uise color Oil a given displzty has beven made and that it -will

j ~bc used in conjunction with other codes (v. g. . :dphanrrrrierics, shapes,

Position. Orientation. etc. )

vhe most difficult and perhaps least. understood aispect of color usc is the

aipplicationl of color as a coding toot when applied to a particular display~

format. Other issues such as synibol si,-e anid luminamnce are better doctl

merited. The data and recommenedation.- are based onl information aivailab] et

in published articl, s and research reports. Ani attempt hias been made to

j use the available dat a as much as possible in the development of guidelines.

The document is divided into two makjor parts. I tart I pr wzits principlesfor color use plus,- suppoi ting dnita where such &0t a exiht . fire principles" of1

IPart I are greneral and canl be applied to virtually) any color display.1

iii

l1,111 11 applies tile pr-inciples in Pavt I to reval- world cockpit displavsbae

Oil anlvis pilot ts display, ulsage ill sever~al seirected high -workload miissýions,

The design peirnciples areu then uised to goeiierae rccouir-endutiozrs for- color,

coding these displays. Examples of display fo-rmats incor~pora-,ting the r-ec-I

ommended colorý coding arc also pr-ovided.

l'a-'t I was wr-itten by AMarj oi-e Ktrebs * Par-t 11 was wrjittenl b), Tirames' Wolt

aind J ohn Sandvig.

'riie contents of thiis documegit arev baised on1 infolniation curr,1ently avaibbit'e.

No experimental validation of conclusions is; made. In some' casesý inf.'r~ent-es

are made about tr-ends, and relationships that have riot been complctel~y vali-

dated experimient ally. It iý hopud] that as miore infor-mation becomles- a\'ailnble

this design guide will be r-evised and improved t~o waike it moret r'elw'.anlt,

.sound, and useful to tile color display designer.

Ac knowledgenient is made of thle advice and suppor-t of i nmilber- at Nzv

personnel who revit wed tile manuy výr-sions of thle design guide a ad imade

hfý" 1 thsirsu st1i apt :zio0n. t~m' i-v ýion. Thos who r-ev jewed thle docu-

ment s or helped in other, ways include: Cindi'. Donýiald II ulsoi, tchinwi cl

monitor-, and Dr. J ohn O'Hlare, both of tile Office of Naival !6esear~chi; R~on

Erickson and Dan W\agnet. of the Niival Weaý'pons CenI er, Ch ira' La11e,

Califovnia; I iovd Hilt chcock and Blil Mulilev of tire Natval Airý 1 )evlop-

woent C'enteri, War-niin.t er, mien l~iIi1ad ('mdi'. P aul Cii t t elii of

NAVAIR.

Finally.~v \%%c would like' to ak '(%%-",dgp tlic ttecit~ical help and giiidatncc

pr-O\v(Cid byV 111V 1L~' 1)1. Wairren It. Te'ich. er of Nv%% Mexi co !ýlatv University.

Dr. rT eich~t wr nlsc studied color codit:,,4 undvr ONiP spon~sorship anid prcvided

sl&slandtai viuidanlce througli lts1 ptibli '-ihd r.-pcwts und Pers~onal. Commuitnication

ovrtepttsv:4

Iv

CONTENTS ¶

PART I. PRINCIPLES ANI) GUIDELINESFOR COLOR CODING DISPLAYS

Section Page

I PART I INTRODUCTION AND OVERVIEW 1

if S'rEpS IN DESIGNING AN ELECTRONIC COLORDISPLAY FORMAT 3

III PRINCIPLES AND GUIDELINES FOR THE EFFECTIVEUSE OF COLOR AS A CODE 9

Phiysical Specifications for Color Symbols 9

Symbol Size and Resolution Requirements 9

Color Display Luminance and ContrastRequirements 15

Display Location and Peripheral Vision 23

Selecting Specific Colors 30

Color Coding Principles 44

Benefits of Color vs. Other Codes 44

Color Used in Conjunction with Other Codes 47

Color as an Irrelevant Coding Dimension 52

Effects of Displayed Symbol Density 58

Coding of Multiple Display Sets 61

vii

PtCBUMQ PA-I WLANK

CONT EN'TS (continued)

PART IT. API1IACATION OF COLOR tI' NCIILt'LS TO

Section FIGHTER/ATTACK AIRCRAFT DISPLAYS Page

I,1) PART II INTRODUCTION AND OVE-RVIEW 6:

V ELECTRONIC DISPLAYS IN FIGHTER/ATTACKAIRCRAFT 64

Aircraft Display Trends 64

Advanced Integrated Display System 66

AIDS Concept Description 68

Display Media and Image Generation 74

Ambient Conditions of Use 75

VI DISPLAY USAGE ANALYSIS 77

Mission Segments 77

A-7 Ground Attack Segment 78

F-18 Ingress and Intercept Segments 79

VFA Ground and Nir Attack Segments 79

Display Usage Taxonomy 80

Display Use Frequency

Link Analysis 102

Conclusions of Display Usage Analysis 107

viii

CONT I,.'NTS (concluded)

Section Page

VI1 EVAtIAATION OF "'ll ADEOFFS IN COLOR

A PLAC AT A ON / 10

Evaluation Vramework 110

Prescribed Requirements 112

Natural E'vironment 113

Systlen Concept 114

Infoi-m'tion Content 119

Mission Operations 124

VIII SUMMARY OF CODING RECOMMENDATIONS WITH

SAMPLE APPLICATIONS 134

l ecormended Display Color Codes 134

Alternative Coding Schemes 135

Constraints 136

Applicable Standards 136

Consistency in Color Code Application 137

Projected MAaps 137

Sample Color Code Ap'lications 137

Electronic Displays 137

Electromechanical Displays 149

REFERENCES AND BIBLIOGRAPHIY 153

APPENDIX A. DEFINITIONS OF TERMS AND

CONC EPTS 175

ix

L.IST OF IIUSTIIATIONS

Figure P'age

1 Rlocommcndcd Symbol Size as a Function of Numberof Colors Used on the l)isplay (Below 21 minutes of arc, )color peirception may be adversely affectcd. ) 11

2 Acuity as a Function of Target and Background Color(The target coisissted of an opening in a Landolt rin-g.Aperture sizes vai jed from 1. 6 to 3. 9 minutes of arc. ) 14

3 Effects of Interaction Between Color and BrightnessContrast with I)irection of Contrast oil leading Time 17

4 Effects of Target-to-Background Contrast Valuc onl'xading Time 17

5 Median Response Time in Milliseconids as a Uunction ofSignal \Vavelength for Five Level- of Symbol I uniianmce(Brackets indicate 0. 01 confidence interval. ) 21

6 Response Time as a Function of W\ avelength 22

7 Be,;ponse Tirnc as a Function of Sional I~uminance 23

8 Retinal Iso-F1T Zones for \VWhitc 2

9 Retinal Iso-WT Zones for Yellow

10 Retinal Iso-l1T Zones for Green 2'7

11 Retinal Iso-UT Zones for B1lue 27

12 Retinal Iso-UT Zones for lied 28V

13 IPcreentage of the Binocular Visual Field IRepresetedby the Iso-RT Zones Presented in Figures 8-12 for EachColor 28

ix

I1

11SfoF'i.IASV3 V (N (contirued)

F'igurec "Ine

14 SUPt'rpositioi) Of iso- 1(1 ZoneCs for fýled upon Siulaltltor(Cockpit Int .,rumcnt P anel and Rlunway Scene 20

15 Illu-stration of P eripheral Color I islv oblem. 1nda Ilropos ed Sýolutioni 3 1

16 Reaiction Timew ts a Function of Practice for. Four ('odeSb n 7 SWith EqUiprobaible Alternatives 3 5

1 7 1 :ff(,ct of Number of ('ode Levels On OperatorI er-forumnee 3 5

18) Performance TIn B eadinig ('01cr'-Coded AlphaInUmerics

:1 ) .1"Fvt iolt)O! of SiZel ='d; Colr4

19 Beading Accurac~y asý a Function of Symbol C7olor- 40

2)0 Aviaition Colors from M11 7,-CK2S03O)A 42

21 B epre~senlarive Advanced Inter-Ated Display Ssek*.lI)S) Formazts 155

22 Relative EfJfects of Task D~ifficulty On Per-formf-ance ofSimulated Pilot inc TIasks5 -s a F,,to,, 1)ft u

Methods of Color Coding 57

2 3 Effect Of I enisity and Displayv E-xposure Time On Accuraic.v 6i0

24 Effect of Color Codingý as a [unction of D~i:play- Dens-ity C60

2 5 Counting Errors as a Function of Displa 'v Densi-ty.('onuparing Color Coding with the Three Shape Codes 6 1

20 General Layoit of E'lectronic Displays in ThreeIlepresent ativo Aircraft 65

xi

LIST OF ILLUSTRATIONS (concluded)

Figure Page

27 HUD Format, Takeoff/Navigation Mode 72

28 HUD Format, Terrain Following Mode 72

29 HUD Format, Guided Weapons Mode 73

30 HUD Format, Landing Mode 73

31 F-14 VSD--Cornbined Stroke-Written and In-RasterImage Generation 76

32 A-7 Close Air Support--Display Usage Taxonomy 839

33 F-18 Ingress---Display Usage 'laxonomy 84

34 F-18 Medium Range Intercept--Display Usage Taxonomy 85

35 VFA Close Air Support--Display Usage Taxonomy 87

36 Deck Launched Intercept--Display Usage Taxonomy 89

37 Representative A-7 HUD Format with and withoutThree-Color Coding Applied 139

38 Representative F-18 Stores Management Formats with

Green and Blue Coding Applied 141

39 Representative AIDS Formats with Three-Color andDesaturated Green Coding Applied 143

40 Representative AIDS Formats with Three-Color andDesaturated Orange Coding Applied 145

41 Representative AIDS Formats with Three-Color,Desaturated Orange, and Blue Coding Applied 147

42 Representative Electromechanical Indicator with Three-

Color Coding Applied 151

IX

xli

LIST OF TABLES

Table Page

1 Recommended Minimum Alphanumeric CharacterHeight for Colored Symbols on High and Low LuminanceDisplays 12

2 Effect of Some Varieties of Colored Light on SomeColored Objects 20

3 Criteria for Determining if a Display is Foveal orPeripheral 24

4 Errors of Color Identification 33

5 Ten Colors that can be Identified Correctly Nearly100 Percent of the Time Under Good Viewing Conditions 37

6 Recommended Colors for a Six-Color Code 38

7 Range of Percent Difference Scores for Several Usesof Color Coding 45

8 Discrimination Accuracy for the Three Single and theFour Multidimensional Symbol Codes 50

9 AIDS Display Functions 69

10 A-7 Close Air Support, Display Use Frequency 94

11 F-18 Ingress, Display Use Frequency 95

12 F-18 Medium Range Intercept, Display Use Frequency 96

13 yFA Close Air Support, Display Use Frequency 97

14 VFA Deck Launched Intercept, Display Use Frequency 98

15 A-7 Close Air Support, Link Analysis 103

xiii

IiLIST OF TABLES (concluded) 1,

Table Page

16 VFA Close Air Support, Link Analysis 1.04 117 VFA Deck Launched Intercept, Link Analysis 105

xiv7t.A

5.a

I-

"PART I

PRINCIPLES AND GUIDELINESFOR COLOR CODING DISPLAYS

Prepared by

M. J. Krebs

I.r

t-4

SECTION I

PART I INTRODUCTION AND OVERVIEW

The question of whether or not to use color in various display applications

is currently one involving some controversy. Data that can support either

side of the issue can be selected from the literature. A careful review

and analysis of the elor literature reveals that the issue of color utility

is not a simple one. The value of color as a coding method is entirely

dependent on its effective use in a specific application. That is, it can be

beneficial, neutral, or distracting. Which of these outcomes will occur

environment, the display medium, and the specific way in which color coding

is applied are all important.

When this project was first begun, it was expected that specific display

formats using specific color codes could be provided in ti-ls document,

allowing the user to select that particular format closest to one of concern.

Unfortunately, this has proved to be much more complicated than was

initially assumed. The "ideal" color display is dependent upon so many

situation-specific factors that a good color display in one application may,

in fact, be a poor one in another. Thus the objective of Part I of the

Color Display Design Guide is to provide the user with data and principles

that will be helpful in making color format decisions. After carefully

analyzing the application in question, these principles should provide the

basis for making correct color coding decisions.

In Part I, R set of guidelines or principles are provided for color use.

Supporting data from the literature are provided in conjunction with these

statements. For the general reader who may not be familiar with certain

concepts and terms in the human vision, display, and color areas, selected

terms and concepts have been defined. These definition3 are provided in

ppendix A. This information has been placed in the appendix to provide I

easy access to specific terms while not interfering with the continuity of

Section III.

2I ,

SECTION II

STEPS IN DESIGNING AN

ELECTRONIC COLOR DISPLAY FORMAT

The effective use of color in any given situation requires an analysis of the

specific environment, the displayed information, and the operator's task-

related information requirements. The following steps are provided as a

guide for the display designer in performing this analysis. The order in

which issues are raised in this section follows to some extent the order of

topic areas and principles presented in Section III.

1. Determine the colors available with the display system

hardware.

2. Determine the maximum luminance achievable with each

color.

3. Consider the ambient illumination in which the display

will be used:

a) Dark

b) Average room luminance

c) Variable--dark to bright sunlight

4. Calculate the luminance contrast achievable for each color

under the worst possible operating conditions (i. e., high

ambient light, such as bright sunlight shining on the display

surface).

3

Any color that will not provide adequate contrast under"worst case" conditions should not be used as a primary

(nonredundant) information source. It should be used

elsewhere with caution.

5. Of the remaining colors available, select up to five

(maximum). The particular colors chosen should be

widely spaced in wavelength from one another.

6. Determine where the display will most likely be located

relative to the operator's normal viewing position.

a) If the display is to be peripheral to theline of sight,

any signal drawing attention to it should be white if

the signal is also peripheral. If colored, the signal

should be placed in the line of sight.

b) If the display is within the operator's normal scan

pattern, it can be considered foveal. Caution should

be taken to determine if this is a correct assumption.

7. The following size constraints should be observed:

a) On the display itself all colored alphanumerics and

symbols should be at least 21 minutes of arc high.

Lines should be about three to four minutes of arc

wide for any graphics.

b) Avoid the use of blue in the coding of alphanumerics

or any small symbols.

i4 1

8. Consider the use of color on the display:

a) When symbols are difficult to see, as when they are

superimposed on imagery, use color as a redundant

dimension to improve symbol visibility.

b) As a general rule use red, gr .en, and yellow accord-

ing to the conventional meanings only (i. e., red =

danger, yellow = caution, green = safe).

c) For data displays use red, yellow, or green alphanumerics

as a partially redundant code to indicate the present

relative status of the numbers presented. For example,

a digital representation of altitude might be coded as

red if it is too high, yellow if borderline, and green if

it is within tolerance.

d) Use color to group spatially separated but related

information (e. g., a series of checkpoints on a map

or friendly vs. enemy installations).

e) Use color to reduce the effective density of items on

a cluttered display by separating them into several color

categories where the symbols can be assigned to task-

related groups.

9. Consider the other displays the operator will be using in

conjunction with this display. If any of them are color dis-

plays observe the following:

a) Similar colors should have the same or similar meanings

across the display set. They should never have contra-

dictory meanings.

5

b) Color can be used to visunlly gr )up information

across displays as well as within one display.

10. Consider the operator's workload during display use. The

higher the workload, the more important the clarity of the

displayed information. Under high workload conditions:

a) Use fewer than the rmaximum number of colors. The

more complex the color code the more difficult it will

be to use.

b) Use color primarily as a fully redundant or partially

redundant dimension (i. e. , to enhance symbol visibility

or to convey relative information quickly).

c) Avoid the use of irrelevwit color (i.e.. a nIlti colored

di,'play where color has no specific, necessary, or useful

task-related neaning).

11. In a11 situations except those of very low workload avoid

the temptation to overuse color. The more colors used

in any one display set the less effective each will he.

a) A colored warning light will lose its attention-getting

power in a panel full of multicolored lights.

b) The most effective color display is one in which color

is used sparingly, only when needed, and where it

uniquely conveys information that other codes cannot

or do not provide.

I!

c) Remember that each color added to the display not

only adds to the potential difficulty of display use, but

also increases the deniands on the hardware for precise

color production.

12. Consider the possibility of display failure. Is the backup.

display a color display? If not, all critical information should

be fully redundant with an achromatic code such as alphanum-

crics or symbol shape.

13. After reviewing the above points, consider the information to

be presented on the display.

a) W1h1t other codes are being used in addition to color?

b) Is the display cluttered?

c) How can color serve to quickly convey information to

the user?

d) How can color uniquely provide information that other

codes cannot convey?

0) If color were n,'. available, what are some of the

major potential problems in display use?

f) I-low can color solve any or all of these problems?

g) What potential Orawbacks can you see in using color

in certain ways (e. g., increased symbol size)?

14. Consider the range of missions or tasks in which the display

will be used.

7

15. Sketch out at least two alternative color formats.

16. Work through the various changes in format for the display

as a function of changes in mission to see if contradictory

or confusing uses of color arlse. Be careful to consider jrelated displays in this analysis. Consider also the

possibility of display failure (Stcp 12).

17. if both preliminary formats appear equally good, determine

the preference of potential users.

18. If there is no other basis for selecting aetween two or more

equally ''good" alternatives, choos-, the one with the fewestcolor uses as calculated over the entire range of missions

in which the display will be operational.

Theýse steps and the data and principles that support them are further

explained in Section 1Il. By considering the questions raised here the

reader will form the basis for interpreting Section III contents with a

particular application in mind.

1

SECTION III

RI1lNCIVIJ'IS AN1 (GUIID)ELINE'S FOR THEEFFECTIVE U'Si': OF COLOR AS A CODE

In this section principles and guidelines are ,rovided for the use of color

as a display code. For each topic area, relevant literature is cited, tables

and graphs shom~ing trends ,nd relationships are provided, and where

possible, key points are sum61.arized in a box at the beginning or end of

each subsection.

F, Irst 1til' Nisicai specifications for color symbols (size, contrast, lumi-

11ance, resolution) are giveii and the effects of ambient illumination are

shown. After this, guidelines for the use of color as a code are provided.

Principles are stressed that are considered important in the effective use

of color to conv-ey information.

PHYSICAL SPECIFCATIONS FOR COLOR SYMBOLS

Svnmbol Size and Resolution Requirements

An important distinction must be kept in mind wvien specifying symbol stzes

for color displays. The difference between seeing a symbol and perceiving

its color leads to different requirements. That is, a sYmbol may be seen

and even identified, but not be large enough for its coloi- to be recognized.

In the following paragraphs size requirements are given for color perception

! m ! | | | | | | | | | | | | | | | | | | 1

I

unless otherwise spe-ified. R equiremcnts will :ilso vrv with syilbol

luminance and contr,-s, and' will be :iffected b, the- rail-e of C.xpectcd

variations in ambient illumination.

SUMMARY OF COLOR SYMBOL SIZE

REOUIREMENTS FOR CRT DISPLAYS

* Aiphanumerics: 21 minutes of arc minimum height

* As the number of colors increases from 2 to 6,mimimum height increases up to about 45minutes of arc

v Symbol stroke widih: two mini'tes of arc minim.n,

a Line width for graph;cs: four minjtes of arc minimum

a Symbol aspect ratio: 5:7 or 2:3 .vidth/height

Syl-nuol Size- -As shown ill Figure 1. ninimunm symbol size required for

adequate color perception x":.iries from about 21 to 45 niinutes of arc,

depending on thc n,,ber of colors sed, .. . 1 If luminance contrast is low or

the display is degraded by noise and/or poor resolution, ;vmbol site sho:0ld

be increased beyond the mininmumn recommended levels.

1 Haeusing, M. , "Color Coding of Information on El'lectronic 1)isplays,Proceedings of the Sixth Congress of the International ErgonomicsAssociation, 1976, pp. 210-217.

10I i

The lower size limit is well documented.12 The consequence of using smal-

ler symbols may be either:

1. The symbol appears to be achromatic (white or grey), or

2. Two symbols of similar color may be confused (e.g.,

yellow and orange).

The latter point is the major reason for increasing symbol size as the nu.1-

ber of different colors is increased.

60

5 0

40

I-

-. 30

220

• 10

O Li L WI I

0 1 2 3 4 5 6

NUMBER OF COLORS

Figure 1. Recommendea Symbol Size as a Function of Number of Colors

UJsed on the Display (Below 21 minutes of arc, color percep-tion may be adversely affected.)

2 Bishop, H. P. and M. N. Crook, "Absolute Identification of Color for

Targets !--resented Against White and Colored Backgrounds, " Report No.

WADD TR 60-611, Wright-Patterson AFB1. OH, March 1961.

! I I I I I I I I I I I I I I I Ii11

Size vs. Symbol Lumivance vs. Information Type--The particular informa-

tion being displayed and the symboý luminance will also influence the

recommended size of colored symbols. In Table 1,3 size recommendations

are provided for three classes of information at two levels of symbol

luminance. Size ranges are expressed as the ratio of syn-bol size to viewing

distance. For a giver viewing distance, symbol height can he determined

uy multiplying this distance by the appropriate table value. Note that as

signal luminance is decreased symbol size must be incre.ised. The type

and "criticality" of the inf-)rmation also influences symbol size. The datv

in 0his table are calculated from achromatic data and have been adjusted to

reflect the increased size requirements for color symbols.

'rABLE' 1. RIECOMlI\NDi:D MINIMTUM ALPIIANUMEIRIC CIIARACTERcnr............'r11 (2 LOIR iZi; ) S1 A\I 1)LS ON IIMGhI AND LO.',' AtM-INANCE DIS PLA YS

Type of - - I i pl1x h 1 ,ow I)is,-plyInformation ,uni Iccw .) umna'nc ?

(ti;,playcd (d /ill() (to U. I 'd /inll)

Critical P ita, 0. 0. 011 Lo 7Variabilc Positio 0,

C('ritico o (. 005 to O. 01l 0. 00 ,8 to 0.017F~ix":ld 1'o0iii on '

Non-ritical J)Data 0. 00V to 0.011 0. 0VA to 0. (0 11

(Chlaracter (eight is cxpruesd :Isa a fraction of vixewingt (oist,,('1c.)

3 Smith, S. L. , "Visual D!, plays - Large and Small, i\I.TU 1K Corporation,for 1;SAF E.lectronic Systems Division, 1.SD-TDd-62-:339, AD 293-826,1,62.

12-_

IJ

Acuity as a Function of Color--The ability of thle observer to discriniinote

fine detail varies as a function of both symbol color and background color.4

In Figure 2, reading accuracy is compared for four red and blue target -

background combinations. The percentage of correct responses for the

various-size openings in a I.,andolt ring (i.e., target detail) is plotted. The

relative performance superiority obtained with red targets is clearly seen

in this figure. The observer is more sensitive to fine detail at the red

versus the blue end of the spectrum.

RESOLUTION REQUIREMENTS FORMATRIX DISPLAYS

* Use larger dot format rather than smaller brighter dots5

* A L x 7 dot matrix will provide marginal performance;larger matrix sizes should be used where possible

RASTER SCAN DISPLAYS

a Fifteen scan lines per symbol height (minimum)

4 Myers, W. S., "Accommodation Effects in Multicolor Displays," AirFox-cc Flight Dynamics Laboratory, Wright-P3atterson Air Force Base,Oil, AFFI)L-T11-67-161, Al) 826-134, December 19C,7.

Ellis, 13., G. J. Blurrell. J. 11. Wa:,-f, and T. 1). F. IHarokins, "The Formatand Color of Small Matrix I)isplays for Use in High Ambient Illumination, "Royal Aircraft Establishment Technical Report 75048, March 1975.

13

RED ON RED

........... BLUE ON BLUE

.------ BtUO ON RED

RU) ON BLUIL

100 I--

90 "

nU) 70

/'- *..'f/ K1/210

LO

L•i- 60

50 1

10

0 1.6.SMALL SI? 1&" 0 APIRTIR{ L ARGE(AR(CMIN)

Figuro 2. Acuity as it Functiton• of 'l'irget t id Ba~ckground Color (Thetarget con.,Asted of an optnhng In it l,:mdolt ring. Aperture'•

,,4zt's vari'hd fr'om 1. (0 to :1.9 mimilte's of treft.) ,

14

Resolution Requirements for Color Symbols--The display requirements for

producing color symbols are closely related to color acuity. If the display

is a raster scan CRT the resolution is typically defined in lines per symbol

height. If it is a matrix or LED display the requirements are defined in

terms of number of dots or strokes per character.

For raste" displays resolution is expressed as the minimum number of

lines per symbol height. A well-established standard for black and white

TV systems is ten lines per symbol height for 100 percent accuracy in6,7.

character recognition. 6 Since recommended symbol size for colored

symbols is about 50 percent greater than that for black and white symbols,

a reasonable standard would be 15 lines per symbol height as a minimum.

Color Display Luminance and Contrast Requirements

The specification of required color symbol luminance depends on a number

of factors. The most important of these are background luminance, ambient

illumination, and symbol size. At very low symbol luminances or under very

high ambient lighting conditions, the color of the symbol is also important.

To specify luminances for i particular application, the entire range of

ambient lighting conditions in which the display will be used must be

specified.

6 Ericksen, C.W., "Multidimensional Stimulus Differences and Accuracy ofDiscrimination, " Wright Air Development Center, Wright-Patterson AirForce Base, OH, WADC TR54-165, 1954.

7Shurtleff, D. A,, "Design Problems in Visual Displays, Part II: Factorsin the Legibility of Televised Displays. " The MITRE Corporation, Report

No. ESD-TR-66-299, AD 640-571, September 1966.

15

SUMMARY OF LUMINANCEAND CONTRAST REQUIREMENTS

Symbol luminance:

* Minimum for good color perception: about 3 cd/rn2

* Optimum under mod rate lighting conditions: rangefrom 30 to 300 cd/ml

Background luminance:

* Visibility of color symbols better on dark background

Contrast:

* For CRT displays, symbol -to-background luminance ratios

of about 10:1 optimum

Ambient illumination:

1 ThhIa~ghe die ambient illumination, the higher the

Ssymbol luminance must 6,- to achieve adequate contrast

Contrast -- Available data derno.'itrate thiat slighitly better v-isibility is

achiieved onl color displays if tile ýsynbols are displayed onl a dark backiground.

Thie revcrsc is true for blick and white displays. This 1'clationsIJip is shown

in Figure 3 8for performiance on a dial reading taskc. In Figure 4 8a perfor-

mance comnparison between color -aind b)lack and whiite s~vmb)c)1 wih qual

lumninance contrast is shown. In this figure thie relativ.e superioritY of' color

symrbols (i. e. , faster reading timeC) is attributed to the additional beliefi!3 of

color contrast. Thi~s advantage is onlyv demionstrated foLý meldiuml lu0)i1anlce

contrast vailues. Analysis of these data indicated that beyond 15 lper.cL'nt coln-

t rast, color sym-bols were not significantly affected. Achromiatic synibols

-were affected at bothi the lowest and] highest values.

NMclxeai, 1\1. V., "firighitness Contrast, Color Contrast, Lind I ~egibilitv, '

I funian Factors5 7, I)eceinber 1965, pr?. 521 -526.

E COLOR CONTRAST

75 BRIGHTNESS CONTRASTo 75

o

~70

LD65H

DARK ON LIGHT LIGHT ON DARK(NEGATIVE CONTRAST) (POSITIVE CONTRAST)

tigure 3. Effectbu, of . .i.eractio• ,• etwen Co••(.hlnr and Brightness

Contrast with Direction of Contrast on Reading Time

PERCENT CONTRAST IS COMPUTED AS FOLLOWS: BRIGHTNESS CONTRAST

L LT _. COLOR CONTRASTSx 100 C 80

WHERE ~75LB B 2ACKCROUND LI MTN.NrFE

LT = TARGET LUMINANCE 7

65

0 15 20 26 32 40 49

CONTRAST VALUE (PERCENT)

Figure 4. Effects of Target-to--Background Contrast Valueon) Reading Time

1 "'1

4

HaeusingI recommends luminance ratios (, 1 /L 2) of about 10:1 for multicolor

CRT displays. Ileglin9 recommends minimum luminance ratios of 5:1 or

8:1. The most relevant ratio would depend on other factors such as symbol

size and number of colors used. If one factor is by necessity at a minimum

value, values on the other factors should be adjusted to compensate. For jexample if luminance ratios of 5:1 are likely, then symbol size should be

increased well beyond 21 minutes of arc, and/or the number of colors used

should be redueed.

Ambient Illuminnation- -For many real-world display applications, the major

factor influencing display visibility is the ambient illumination. When

external lightinrg can be maintained at a minimal level, achieving adequate

visibility is relatively easy. When the ambient lighting is variable and/or

becomes very bright at times (such as is found with cockpit displays in

bright sunlighkt), problems arise. The effect of adding environmental light

to a display surface is to decrease the symbol-to-background contrast.

Colors begin to desaturate or fade, and under very high ambient lighting

they may be completely washed out. Conversely, if outside lighting is very

low it ma', be dcsirable to k eep symbol luminance at a minimunm to maintain

the operator's adaptation to the dark. If the symbols are colored, reduction9

of their luminance below about 3 cd/ir will seriously interfere with the

perception of their color. If either of tlhese situations is likel'y to occur

across the range of expected display uses. color should probably be u.ed

as a fully redundant code.

9 Heglin, J. , NAVSIIIPS Display Illumination Design Guide, Section 11.: hu1man

Factors. Naval Electronics Laboratory Center: San Picgo, CA, 1973. j

Low Ambient Illumination- -Perception of surface colors on maps, charts.2

etc. requires luminance values of at least 3 cd/rn . Below this minimum

level it becomes difficult to differentiate colors. Comfortable reading and2

good color perception require from 30 to 300 cd/m . At very high lumin-

ances (beyond about 3000 cd/m2 ), surface colors bccome increasingly hard

to see due to poor luminance contrast.

If the ambient lighting itself is colored, (such as the red night lighting in

some aircraft), surface colors become more difficult to discriminate. They

may markedly change in appearance. Table 210 describes some of these

changes.

Both th'2 intensity and the color of the environmental lighting bhould be

cc .sidered when choosing specific colors. Particularly for surface colors,

the colored ambient lighting may cause color confusion.

When the display is to be used in light-restr' cted or night time conditions

and the background is dark (below about 1 cd/m2 ), required symbol

luminance is lowest,

In Figure 5,1 observer response time data are given as a function of signal12

luminance and woAvelength. Below about 0. 1 cd/in the symbols are not

1 0Semple, C.A. Jr., R.J. Heapy, E.J. Conway, Jr., and KT. Burnette,

"Analysis of Human Factors Data for Electronic Flight Display Systems."

Technical Report No. AFFDL-TR-70-174, April 1971, 570 pp.1 1 Pollack, J. D. . "Reaction Time to Different Wavelengths at Various

Luminances, " Perception and Psychophysics, 3, 1968, pp, 17-24.

19

TABLE 2. EFFECT OF SOME VARIETIES OF COLOREDLGIIrr ON SOME COLORED OBJECTS

(.lTI ('"(I 01O.U E 1 ) 1.l6 1rl 111.t 1 I .II11 (;P! I N 1.11 N 1.1,(W\i I. fli )' h , 1.igh pink V ,r\ 1i h Li lu, , V ,r\ Ii .l rr'i t 'n V ,rN l igit \,' l1

;,' d Irilli , rd1 i h r) rlh o i- h i 'd ltiigh ri'd

I ii'ht 1.-juc |P,'ddl.-h I Wit" }•tr lgh, lt " 1.r.1.1v} L.•,,[ight I-edtil'1h h~imI

U.i-k L]. l ,r k re ddi sh pu rple 1•1 "zr 1 .nIt ,I,'l , I'.,rk gr• -• it.h i ,.-e l.i J0 0 r 'd di,-l I pu lp ,

1. ixh 1 I ro -ir i , I ix'. hlig gli , ,hi N. i " III" xi.,11 t ligrhtge

seen as achromatic rather than colored signaLs. In these conditions, shorter

wavelength signals in the blue to green region produce much faster response

times than do the longest wavelengths toward the red end of the spectrum.

High Ambient Illumination--In general, the effect of ambient illumination

striking the display surface is to reduce the symbol-to-background contrast.

Under very high levels of ambient illumination, response time to signals at

both the red and blue end of the spectrum are faster than those in the yellow12

to yellow-orange region. In Figure 6,1 response time as a functicn of

12Tyte, R., J. Wharf, and B. Ellis, "Visual Response Times in High Ambient

Illumination, " Society of Information Displays Digest, 1975, pp. 98-99.

20S|5;

-4

.-.)X

Q)l

C.)

CIQ Cý

'-'0 0

Lil U

CO

0.0

-~CC

E 1 0E

U') U) " IC) ko

r-~ CD CDC

(03SW) 3Wi3S0) Ni~

21

2

BI--

(AA

LU

•0. 2 II

450 500 550 600 650

TARGET WAVELENGTH (nm)

LUMINANCES: A = 48 cd/mr2, B = 30 cd/rn2

AMBIENT ILLUMINATIOIN: 105 lumens/mr2

Figure 6. Response Time as a Function of Wavelength

signal wavelength is shown at two signal luminances under an ambient52

illumination of 10 lumens/mr "The lower signal luminance produced a

much greater effect (increased response time) in the yellow region. At

105 lumens/m2 red symbols are more visible than green symbols. To be

equally visible in high ambient light, green signals should be about three

times the luminance of the red.

12In Figure 7 , response time as a function of signal luminance is plotted

for red, green, and yellow signals. In this figure the marked superiority jof red under high ambient conditions is demonstrated. Ii

12 Tyte, R,, J. Wharf, and B. F'lis, "Visual Response Times in High Ambient

Illumination, " Society of Information Displays Digest, 1975. pp. 98-99. j

2

RED GREEN

w - YELLOWLU

,. 1.0

0. 1

10 50 100 5002

SIGNAL LUMINANCE (cd/rn)

AMBIENT ILLUMINATION: 105 lumens/m2

Figure 7. Response Time as a Function of Signal Luminance

Display Location and Peripheral Vision

The eye is most sensitive to color within only a small part of the total field

of view. Beyond this area the eye is differentially sensitive to individual

colors, both in terms of limits of field of view and in operator response

time to different colors. (See Table 3 for a distinction between foveal

and peripheral displays.)

Peripheral Sensitivity to Color--Of the various colors the field of view is

widest for yellow and narrowest for red and green (see Figures 12 and 13).

Within the total field of view, response time to different colored signals

also varies. Reaction times (RT) obtained for red, green, blue, yell3w,

23

El?

C).

CUi

Cii-o

ELI,CA)

o CUt-

ri

C) C) C cu

c C'- u 1 .- "U t0

UI vo ) ( 0 C) 1

24J C

and white were mapped over the visual field in one study.13 These RTs

fell into bands. When plotted as in Figures 8 through 12, these bands of

similar response time ("iso-RT" zones) show some distinct color differences.

In these figures, zones are plotted and the related RT within each zone is

indicated. These data are summarized in Figure 13, which indicates that

white has both the widest field of view and the shortest response tinmes over

the entire field, while red has both the narrowest field and the longest

response times.

Blue, green, anud yucliw are roughly similar to each other end fall between

the red and white plots. In Figure 14, the iso-PT plot for red is super-

imposed over a cockpit instrument panel. From these results it can be

concluded that a signal light should be placed as close as possible to the

direct line of sight. White is the best choice for a signal light and red is

the poorest, as indicated by reaction times. The further into the periphery

the light is moved, the greater the discrepancy.

13Haines, R. MW., L. M. Dawson, T. Galvan, and L. M. Reid, "ResponseTime to Colored Stimuli in the Full Visual Field, " Report No. NASATN D-7927, NASA Ames Research Center, Moffett Field, CA, March,1973, pp. 27.

25

I-45

V I IIi

270 ... / .--. . ..--

- , I/ , t

S/

N /

r-ILLD O)F XiWE1INOCI LF/R-hq(

Figure 8. Retinal Iso-RT Zornes •or White

YELLO0

115,

I /,>C•.I '---4 ,< <"- 4

, . , .... .. / ,Vi i

- 35*

HELD OF VIEWTBINOCUL piRMONOCULAR

Figure 9. Retinal Iso-IiT Zones fror Yellow

1*26

It

GREEN

271. 77' 90*

BINOCULARMONOCULAR

Figure 10. Retinal Iso-RT Zones for Green

BLUE

01

270*. ,Y "- 90*

2P'5

FlL fVIEW IUBINOCUL ARMONOCUL AR

Figure 11. Retinal Iso-RT Zones for Blue

2 7

I

VJ

//i~i•'b- <- ,/./

'I' /

Figure 12. Reia _s-R Zoe fo Re1~~ JY

- -------

Figure 12. Retinal Iso-RT Zones for Red sn

LII

Figure 13. 2~ereentage of the B~inocular Visual Field IRepresented by the Iso-T{T Zoncs Presented

in Figures 8-12 for Each Color

28"

I I II l

Peripheral Displays- -If many displays are being used by the operator, some

must by necessity be located peripheral to the line of sight. If these displays

are routinely scanned the problem may not be significant. However, for

those outside the normal scan pattern, special precautions should be taken.

For example, a given display may be located outside the normal scan pattern.

it may contain system status information, which is usually in tolerance. The

information on the display only becomes critical when a tolerance limit is

exceeded. If the display were centrally located, red could be used to alert

the operator. However, red is poor when used as a peripheral cue. What

should be done? Several color coding possibilities exist. The most straight-

forward one would be to use a central master warning light as shown in

Figure 15. Another solution would be to display the warning light or

abbreviated message directly on the primary display.

Selecting Specific Colors

Where color is to be useru as part of a display code the designer must

determine 1) how many different colors will be used, and 2) what these

colors will be.

How Many Colors Should Be Used ?--The decision to use a given number of

colors must be made after considering the limitations of the displky medium,

the ambient lighting, and perceptual limitations of the observer or display

operator. All of these factors are important in arriving at a proper

decision.

I"30

I I I II I I I I II 3I 4

PRIMARY

(STATUS DISPLAY) DISPLAY

1MAL- 1 S[CONDARY SECONDARYFUNCTIONt DISPLAY 3#1REPOýRT,ý

(SQ

SECONDARY" DISPLAY #2

a) Arrows indioate nornal scan of operatoer on car displays. Luf..-display is outside of this pattern. Aessage printed ir red would not helpin initially alerting operator.

PRIMARY

(STATUS DISPLAY) DISPLAY

I- AL-ECONDARYD... SE•CONDARYi"F'NCT!O. SCNDR K<m MASTER "

SREPORT AN.I DSLY#

-'-'----• D IS PLAY #2

b) Addition of mu~ster warning light in center of scar, area provides effec-tive -alerting signal. Directional arrow would further reduce operatorresponse time. Warning light could be colored red.

Figure 15. Illustration of Peripheral Color D~isplay

Problem and a Proposed Solution

31

FACTORS INFLUENCING NUMBER OF COLORS

"* Surface colors vs. self-luminous displays I"* Ambient illumination

" Operator workload

"* Color code relation to operator task IGENERAL RECOMMENDATION FOR

COCKPIT CRT DISPLAYS 1

" No mor6 than four colors on any one display

D• splay .~edium- - Self- lurninous d•p' -- id s ES a o a.... -'-_jroL;Eds y.j sucri as CT~tls may or may

not have a limit on the number of different colors possible. The important

point to remember is that the gr-cater the number of colors used, the

greater the demands on the system for precise reproduction of each color.

The more similar two colors are, the more critical it is that each be pre-

cisely defined and reproduced on the display, The probability of operator

error will increase if color control is not fairly rigid, due to confusion :is

to which color is being displayed. In Table 4,1 an example of some common

color confusions is presented. Note that these confusions occur only for

adjacent colors in the table.

1 4 Connolly, D. W., G. Spanier and F. Champion, "Color Display Evaluation Ifor Air Traffic Control, " fleport No. FAA-RD-75-39, Federal AviationAdministration, Washington, D. C., May 1975, 41 pp. 1

33'2

TABLE 4. ERRORS OF COLOR IDENTIFICATION

Shown Called

Red Orange Yellow Green Total Percent

Red X 21 0 0 21 2.9

Orange 9 X 10 0 19 2.6

Yellow 0 6 X 15 21 2.9

Green 0 0 6 X 6 0.8

Grand Total = 67 2.3

Human Perceptual Limits--With very extensive practice and under15

ideal conditions, human observers can individually identify up to 50 colors.

This number, however, far exceeds any reasonable number for operational

conditions outside of the laboratory. With less practice but still under

1;hboratory conditions, it has been found that as the number of colors increased

the number of identification errors also increased:

Percent ofNumber of Incorrect (Error)

Colors Responses

10 2.512 4.515 5.217 28.6

1 5 Hanes, R.M. and M.V. Rhoades, "Color Identification as a Function ofExtended Practice," Journal of the Optical Society of America, 49, 1959,pp. 1060-1064.

33

It

If the operator task requires absolute identification of a color, ten colors

appears to be maximum for good accuracy. If absolute identification is not

required more colors can be used. Up to 23 colors can be profitably used

in coding of surface color maps where color served as an aid to search,16[

according to the results of one study.

Number of Colors vs. Performance--As the number of clors is in-

creased in a situation where symbol color is assigned a particular meaning,

both error rate and detection time increase. The general relationship

between code size (e.g.. number of colors) and response time is shown in

Figure 16!7" The greatest effect occurs early in training and diminishes

with extended practice. This figure clearly shows, however, that the

greater the number of colors used the more time required to respond to any

individual color when it appear s.

18Similar relationships have been reported by Ilitt, who avetaged the num-

ber of responses per minute over a variety of tasks involving multiple target,,.

This is shown in Figure 17. Stated more positively, when fewer colors are

used the response tim.e .A_1! he faster for P,-h onn -when it nppears.

1 6 Shontz. W.D., G.A. Trumm, and L.G. Williams, "Color Coding forInformation Location," Iuman Factors, 13, 1971, pp. 237-246.

17 Teichner, W.T. and Krebs, M.J., "Laws of Visual Choice Reaction Time,Psychological Review, 81, 1, 1974, pp. 75-98.

18Hitt, W.D., "An Evaluation of Five Different Abstract Coding Methods," IHuman Factors, 3, 1961, pp. 120-130.

34

Parameters:8

1.4 CODE SIZE-4

1.2 3

1.0

122

.8

0i

.6

.4

I 1 I

2 3 4

LOG 1 N NUMBER OF TRIALS

Figure 16. Reaction Time as a Function of Practicefor Four Code Sizes with EquiprobableAlternatives

12

U- Q 10

r 5.

2 4 6 8

NLUMBER OF COLORS

Figure 17. Effect of Number of Code Levels on Operator Performance

35

The importance of increased response time (or increased errors) must be re-

lated to the operator's task. If the workload is expected to be high or fast

reaction time critical, then the number of colors used should be kept as low

as possible.

General Be commenciations- -Although no specific data appear to exist

from real-world displays, several investigators have recommended

that a three-to-four-color limit be used for operational displays. The

smaller code size for operational situations is based on the expectation

that ambient lighting may at times be high, that displ.3 reliability may be

limited, and that fast reaction time of the operator may often be critical.

I •rl•tý &•iors Should He Used ? - -The best general criterion to use in -etecting

a set oi colors of specified size is to pick colors as widely spaced in 1Nave,-

length. ý,s poss'ble along the visible spectrum. Under good viewing ccnditions

!he ter, colors indicated in Table 5 provide a highly identifiable set.

CR!TFRiA FOR SELECTING SPECIFIC COLOR SET

l 0 Maximum waveiength separation

* High color contrast

S.Hi(,h visibility in specific application

* Compatibility of ua with corve-itional meanings

• Legibility an-d ase of reading

* HiGh saturaticv;

19Baker, C.A. and W. I,'. Grether, "Visual l-rese!:•ation Uf Information,Wright Air Development Center, WADG-TR-5416t;, Al) 43-06-1, 1954.

336

I1

TABILE 5. TEN .OLO1S Tt'AT CAiN IIE IDENTI,'ID COIIRECTLYNE'ARLY 100 PE'ICENT OF TIHE TIME UNDER GOOC4DVIEWING CONDITIONS

Domin.a nt ColorW avelength (11m) Name

430 violet476 blu e

494 greenish- blur504 bluish- green

5 15 green556 yellow-green5 82 yellow596 orange610 orange-red642 red

ltowever, as indicated in the preceding subsection, the limitations of

hardware, the effects of ;,inbient illumination, and operator workload

may limit this nurtiber sonmewhat,

Table 6 presents a six-color code recommended by Cook-.2 0 ,,',.: ta.,'

also provides several notations helpful in identifying each color.

0Co-k, T.C., "Color Coding--A IRe iei\- of the Literature," U.S. ArnlyPuman Engineering Laboratory, Aberdeen iProving Gr'owid, MD, 111:1.T•'ch Not(: 9-74, Novemnber" 1974.

37

C.) >

-4 '. '

o1 :1ucn 0 cocnD L

C) C.0 CV

o C ~ 4 -

C*.

C- ,14

C, [It C:>

C) In LO C.

C) C

(5k

Rielative Visibility of Individual Colors- -All colors are not equully

visible. Even if a specific color can be accurately named when simply

presentitd as a test patch under good viewing conditions, in other conditions

or other tasks it may create problems. The best docurnenied example of

this Is the culor blue. 'ile fovei of the human eye, which Is sensitive to

detall. is essentially blue-blind 2 J Aýa (:o1nskcice, smal syn--bols or

fine- dctal 1-re not seen as well in blue. Because of thils, blue is not recurni-

mnended as a color to be used for alphanumerics. 11nes, etc. unless they -ire

unusuallY large. Bvtter yet, blue should be reserved for coling large zones

or :rcas on thea display.

Relative legibility of six colors (including white) as a function of symbol

size is -7hown ir. Figure 18. The drAa 2 show the speed at wvhichi alphanurner-

ics can be read as a function of symbol color and size. Under the Conditions

of this test red, white, and yellow szymbols were read at a much higher rate

than green or blue sy rnhoLi. Similor data 23are shown in Figure 10'.

W ald. (1. , "B~lue B~lindness in the Norimal Fovea, " Journal. of theOpticalSceyof Amierica. 57, 1967, pp. 1289- 1',03,

22 lvMeister, 1). and 1),J. Sullivan, "(uide to THumine Engineerin- , Design forVisual D)isplays, "' Offi~cc of Naval Recsearchi Contract No: N00014-68-C'-02711, Office of Naval Research, Engineering Psychology B1ranch. Washling -ton, DC, Al) 693-2 370, 1 9G9.

2:ijj~y ' - l. , "1)ichroic Filter Specification for Color Addivive Displays. 11.lFurthlic Exploration of Tolerance Areas anid the influenceý of Other D~isplayVnrt:h1les. TISAF Bonme Air lDevelopmneit Center, RADC-T1I-67-513. AD)f'59-344C ',eptexnber 19(67.

391

22 .

20 - MEAN/ALL COLORS/*. .. %.'

18 -.-. YELLOW

u- 16 . . RED/- ........ WHITE

"14 /- MAGENTA-,f 12 - / / CYAN

'"'- 10 /GREEN

BLUEC=

"6S• l I I

17 20 23 26 29

CHARACTER SI7 E (MINUTES OF ARC)

Figure 18, Performance In Reading Color-Coded Alphanumer'cs

as a Function of Size and Color

26

• 24Vr) >_

V) 22

h_ : _,

CIo tor(.~) 1 6 L

2_1~ I z U

SYMBOL COLOR

Figure 19. 1eading Accuracy as a Punction of Symbol Color

I40-

3_

Visibility of Colored Signal Lights- -Amber, violet, and red are spotted

more quickly and named with fewest errors in both laboratory and air-to-24

ground flight tests of eight different signal ligbts. Incandescent lights

were the worst signals by all measures. It was also found that flashing

ground lights were seen no more quickly than steady lights if thc incen-sity

of the two were the same. The attention- getting power of a flashing sthobe-

type light may be due to the relatively greaLter intensity of each flash.

Conventional R, G, Y Color Code--The three colors, red, green, and

yellow, and their associated mcanings of warning, safe or advisory, and

caution, respectively, at e widely known and accepted color conventions.

Color display design should adhere to these conventions where appropriate.

If the type of information being presented on the display can be readily

assigned to these categories, the resulting display format will be easily

learned by the operator. Military Standard 411-D25 lists the use of the

colors indicated above.

In Figure 20, chromaticity specifications for red, green, and yellow are26given, as well as for several other colors. (See Appendix A for an

explanition of the chromaticity diagram.

24 Iilgendorf, 1. L., "Colors for Markers and Signals: Inflight Validation, "

AMRL TE-71-77, Wright-Patterson Air Force Base, OI1, AD 737901, 1971.

" 5 Military Standard, MIL-STD-411r), 30 June 1979 with Notic ;-1, AircrowStation Signil.s, 30 August 1974, Washington, DC: Department of Defense.

2 6 Military Specification, M-L-C-25050A (ASG), 2 December 1963, withAmendmvent-l, 30 September 1971, Colors, Aeronautical Lights andLighting Equipment, General Rcquirem.nts For, Washington DC: 1. S.Government Printing Office.

41

ft I1 2 3 4 7

53 . .UAIANE RGYAT O ,

I jU S AVIATION COtOH LIMITS a

_-_ I!)4

I -1: . .. 41

AVIATION GREEN *0 4.. 0 32'0vi"I

SAVIATIONIWHITE

0 170 AV 9b97'107 YELLO 0

.60 V-0 370i 0 540

Yl 0 "390 0 ,ý ,O 335._

- • .• AVIATION RED,)"

I RPLE

4B I '

w -q'0 175

0. .,- I, I

F'igu~re 20. Aviation Color-; fromn MIL-C-25050A

42

SPECIFIC COLOR RECOMMENDATIONSFOR CRT DISPLAYS

* Use no more than four colors

* Use red, green, and yellow to code alphanumerics

* Use blue for large symbols or where symbolidentification is not a problem

0 Use conventional color code when appropriate

* Use white for peripheral signals

Use of Saturation Difference in Color (:oding--The preceding discussion of

which colors to use on a display has been concerned entirely with differences

in line. The general recommendation would be to use highly saturated color,

(hues) to maximize the differences between colors. In some situations it may

not be possible or desirable to use highly saturated colors alone. One major

reason for adding saturation as another color dimension would be to increase

the number of color steps achievable with a particular display medium. For

example, if the designer had only a two-color display, it might be possible

to use saturation differences to produce nmore than two discririnable steps in

the color code. A red-green display could then become a four-color display

by ý.roducing a high and low saturation version of each color. Thus, red

would bec')me recý and pink, and green would become light and dark green.

Saturation differences awe used to produce the many color variations on maps

and other printed (surface) color materials. Ilue-saturation combinations

can provide a large number of discriminable different values for the color code.

43

Caution should be taken to ensure that Ute changes in saturatioln do not

produce colors that are difficult to see under some viewing conditions. Hligh

ambient illumination on the surface of the display will, itself, tend to

desaturate or wash out the color of a symbol. If thet symbol is already

desaturated its visibility may be seriously degraded. t ap

If ambient lighting can be controlled in those situations where the display is

to be used, saturation level may provide a good method of increasing the

color code size. If lighting will vary widely, use saturation level change

only with caution. The visibility of desuturated colors under all level-' of

expected ambient lighting should be tested prior to use.

COLOR CODING PRINCIPLES

Benefits of Color vs. Other Codes

COLOR CODING WILL BE HELPFUL IF

* The display is unformatted

* Symbo: density is high

* Operator must search for relevant information

* Symbol legibility is degraded

* Color code is logically related to operator's task

44 !.

When used appropriately, color coding can provide significant performance

improvements compared to other codes. it is clear from the available

research literature that the relative effects of color coding are strongly

determined by the specific function being served by the code. Results

obtained from a number of code comparison studies 27are shown in Table 7.

This taole shows that there are situations in which color is clearly superior

to other co~ies. It also shows that in some situations color is significantly

inferior as a code~. One major factor that determines thie relative mierits

of color coding is the task performed by the operator.

TABLE 7. RANGE' OF PEJRCENT DIFFERENCI: SCORES FORSEVERAL USES OF COLOR CODING

2 7 Christ. RI. L. and G. M. Corso, "Color flcsearch for Visuz'l Displays,Technical Report No. ONH-CR2l3-102-3, July 1975, 108 pp.

45

1

Results of two code comparison studies18, 28 demonstrate that color coding

is clearly superior to numbers, shapes, or letters when the task involves

locating targets in a cluttered field of nontargets. Numeric coding is sup-

erior for identification tasks. Both the numeric and color codes are hene-

ficial in a counting task. The remaining tasks (comparing and verifying)

showed no specific advantage for any of the codes,

An investigation of symbol identification time for seven different codes

including numbers and colors was reported. Numeric coding was superior

to all others and color coding was the worst of thp six. These results

demonstrate the fact that color coding when used alone is a poor choice for

a task requiring identification, To provide performance benefits, color is

most effective as an aid in locating target symbols quickly. Its value is

attributable to the added discriminability provided by color.

"28Christner, C.A. and 1.LW. Ray, "An Evaluation of the Effect of Selected

Combinations of Target and Background Coding on Map Reading Performance- -

Experiment V." Human Factors, 3, 1961, pp. 131-146.

2 9 Alliusi, E.A. and P.F. Muller, Jr., "Verbal and Motor Responses toSeven Symbolic Visual Codes: A study in S-l- Compatibility, " Journal ofExperimental Psychology, 55, 1958, pp. 247-254.

46 I

USE COLOR

* To aid operator in locating particular information

* To draw attention to some specific place or symbol

USE ALPHANUMERICS

* To convey specific status information

* To identify specific targets

Color Used in Conjunction With Other Codes

USE MULTIDIMENSIONAL CODING

* To convey specific information t.at cannot otherwise

be conveyed

I Tv innurudsu the amount uf informatiun that can bedisplayed

On any complex display a number of coding dimensions are typically com-

bined to convey specific information. The most frequently used codes are:

alphanumerics, shape, symbol orientation, symbol size, and symbol bright-

ness. Color can be used in combination with any of these codes to provide

additional information, or to make existing Information easier to see or use.

47

If the particular color of a symbol is correlated with a value or values on

another dimension, then color is a redundant dimension. Full redundancy I.

occurs when this correlation is perfect, such that knowing the value on one

dimension completely determines the value on another dimension. If this Icorrelation is not perfect, as is the case when fewer valaes are used on one

of the two or more dimensions, then there is partial redundancy. [

An example should help to clarify the meanings of full and partial redundancy.

A hypothetical digital readout has nine possible values it can assume. If

color were rully redundant with numeric value, then each of the nine digits

would be associated with one of nine different colors. Knowing the color of

the symbol would provide full knowledge of the numeric value and vice versa.

If, however, several numbers were associated with the same color, such

that for example the three lowest values were coded yellow, the three middle

values green, and the three highest values red, then the color code would

be partially redundant with the numeric code. That is, knowing the symbol

color would give only partial information about its numeric value. Knowing

that the symbols displayed were green would indicate that the numeric value

was one of three intermediate values.

A third form of multiple coding involves use of two or more codes in a situa-

tion where each conveys unique information not contained in the other codes.

Such coding is nonredundant.

I j

I

USE FULLY REDUNDANT CODING

* To improve symbol detectability

* To aid in discriminating among symbols

USE PARTIALLY REDUNDANT CODING

" When information can be categorized at more thanone level of specificity

USE NON-REDUNDANT MULT!FLE CODING

"* To increase the tate! ,umntm-r ol Identifiabie catagorioi

j! VNonredundant Use of Multiple Codes--N, onredundant color coding can be

used to increase the number of symbols which can be absolutely iddcktfied

on a display. Several studies have reported inproved ability to dis-

criminate among objects when size and colo"v or size, color, and brightness

were combined in a single display thani when any v.er. used ait'o. (See

Table 8. )

3 0 Garner. W.R. and C.G. Creelnan., "Effect of 1Redundancy and Durationon Absolute Judgments of Visual Stimuli, "Journal of Experimental lss o-og, 67, 1964, pp. 168-172.

49

I

TABLE 8. DISCRIMINATION ACCURACY FOR THE THREE SINGLEAND THE FOUR MULTIDIMENSIONAL SYMBOL CODES

Symbol Code of AbsolutelySymboymbols Discriminable Symbols

Size 7.19

Hue 8.45Brightness 5.06

Size-Hue 11.90Size- Brightness 7.89Hue-Brightness 13.55

Si cO- I4, Pc- • tni Pccz 17.28 1

Practical applications of nonredundant coding include, for example, the

coding of friendly and enemy "targets" on a map, sensor display, or air

traffic controller's display. Targets could be coded by color as either

friend or foe. Further distinctions as to target type (aircraft vs. land

vehicles) could bue cuuud by shape. SpeciL1L. Lf I gts Vwithin1 A Lype cVuLU be

coded alphanumrierically. Using such a system, a large number of targets

could be uniquely coded in such a way that each is absolutely identifiable. 1

Use of Totally Redundant Codes--Symbols may be difficult to discriminate

because the display is degraded by noise or poor luminance contrast, etc.,

or they may be difficult to locate because of clutter. In such situations color Imay be used as a totally redundant dimension to improve symbol discrimina-

bility. Targets that can be defined on several dimensions are found more

quickly than when either dimension is used alone. Color and shape have

50

1

bcen fou:nd to be the best combined code. Other dimensions such as bright-31, 3 2

ness and size were not as effective.

Use of Partially Redundant Codt's---When color is combined with, for example.

an alphanumeric code, such that groups of numbers or letters are similarly

colored and several groupz are defined in term.s of specific colors, then

color is a partially redundant dimension. Such con.bined coding is used to

provide relative status (using color) and sjpccific status (using alpharnumerics).

For example, altitude coui'd b.e coded as high, in-toler.ince, 01' lw. using a

three-color code such as yellow. gree:. and red, respe,'trv,--iy. Actual, also-

lute altitude would be given digitally. Depending Tpon pr esent information

requirements, a quick glance at the display woild irifLrm t i pilot of the rel,- -

tive altitude as compared to plan. Specific information would be avait-ble. iF

required, by the numeric value. Such mult.lfle cadi:,g would be useful only iu

situations where both general and specific status information are meaniotful

at differ'ent times.

Redundancy May Not Always Aid Performance--In some situations, addiag

a redundant dimension may interfere with the effective use of other, odcz.

If the operator has a strategy for using alphanumeric information, for

example, and redundant color is added to the display. the colors la'" eithcr

"311Ericksen, C. W. and II. V, Itake, "Multidimensional Stimulus Differencesand Accuracy of Discrimination, " Journal of E.xperinmental Isychology,50, 1955, pp. 153-P30.

32Saenz. N. E. and C. V. Riche, Jr., "Shape and Color as Dimensions ofa Visual Redundant Code," Human Factors, 16, 1974, pp. 308-313.

51

provide no benefit or may interfere with task performance. This effect

was reported in a study33 in which the operator was to keep track of the value

on each of a large array of digital readouts. Adding redundant color did

not aid performance in any way. Substituting color for digits resulted in

poorer performance. When the strategy, employed by operators was

examined, the reason for such results became clear. Single digital read-

cuts were "chunked" by operators into multidigig numbers. Colors used

alone could not be easily chunked. Unless redundant color coding permits

the use of a new and more effective strategy for information extraction, it

should be avoided or used only if it provides other benefits such as increased

syrribod legibil ity.

Color as an Irrelevant Coding Dimension

WHEN COLOR CODE IS IRRELEVANT-10 OPERATOR'S TASK

0 -C al- or S er v es * 0 tl 'ctru Ct Q fl er2a fl

S Sin,!iariy colored items nway, be visually grouped in

nonimeaningful or distracting ways _.

SC~olor can functionally becom•e "itoise"

33kanarick, A. F. and P-, C, Petersen, "Redundant Color Coding and Keeping-Track Perfornmace, " Hunman F•adtors. 13, 1971, pp•. 183-188.

50

When the color of a dispLayed symbol has no direct bearing on the operator's

task, the color code can serve to distract the operator in performing that34

task. "Signal" can be defined as those aspects of a display that aid the

operaior in locating a target. "Noise" can be defined as those aspects of a

display that detract from locating the target. Any irrelevant code becomes

noise. Following this analogy, irreievant color in a muLiticolor display adds

to tihe noise and reduces the signal-to-noise ratio. The greater this noise,

the poorer the operator performance will be.

If operator information requirements were always the same, the •-pplication

of a color code would be relatively simple. A more typical problem occurs

when the task and the operator information requirements vary while the

display format remains constant. The result is that a code relevant for one

task may be Irrel.evant for another. Therefore, it would be advisable to

use color codihg (or any code) with caution and with a consideration of the

entire range of display uses and applications planned for that display. To

be niaz:jmally effective a given color should Le related to the operator task,

mnd its presence should conv/ey specific information.

Example of l1elevant vs. Irrelcvant Color--An illustration of the distracting

Sffeet of irr2luv:int color is pruvided by the results of an experimental study

coaductedi under •his contract. A simulated set of VFA/VSTOL aircraft

34 Gretrt. B. F. and L. K. Anderson, "Color Coding in a Visual Search Task,"Journal of E'xperimental Psychology, 51. (1), 1956, pp. i!V-24.

53

cockpit displays (see Figure 21 for an illustration of achromatic formats 35)

were color coded in several ways and the effects of the different color codes

on performance were analyzed. The displays were static (projected slides)

and the operator's task was to detect out-of-tolerance conditions in any of the

major parameters displayed. Therefore, the displays simulated cockpit

displays and the operator's task simulated one important piloting function. A

two-dimensional tracking task was added that varied in difficulty from low

to high, simulating another aspect of piloting an aircraft.

The displayed symbology was coded in each of the following ways:

1. Achromatic symbology- -no color2. Three-color displays with color u.to " psimilar

S...... ...I -,= to1 - group si ia

items (i. e., all scales green, all alphanumerics red,

other symbols yellow)

3. Three-color display where color served no apparent

function (i.e., some scales were green, others yellow;

some alphanumerics were green, others red, etc.)

4. Three-color disnlnv ia n -1hih . San .•icalesI -. I........ - 0.,., 1j ,*..'.L ,II dIS U • 3•}

were green or yellow. Red was used solely to indicatc an

out-of-tolerance condition. A symbol or indicator

would be displayed in red in such cases, rather than in

its "normal state" color

3 5 Linton, P.M., "VFA-V/STOL Crew Loading Analysis. "Report No.NADC-75209-40, Crew Systems Department, Naval Air DevelopmentCenter, Warminster. PA, 15 May 1975.

.4

54 I

ý t-

CC3

4ý

Lz'I

¼5 5

i iGiven the operator's task in this experiment (detect out-of-tolerance

parameters), only the fourth coding scheme was task relevant. The second

and third color codes were irrelevant. Both the effects of relevant and of

irrelevant color were of interest.

The results of this study are shown in Figure 22. A combined score reflect-

ing both tracking accuracy and detection speed are shown for each of the

four color codes as a function of tracking task difficulty. As was seen in

Figure 21, when the task is easy (low workload), there is very little

difference among tLe codes used. As the task became more difficult, how-

ever, the effect of the different codes is quite different. The color-as-cue

code is far superior to the others. The color-as-organizer code is no

different from the achromatic code. Note that the co t,,I ,on in which color

has no task-related function yields the worst performance. In this last

case, color served only as a distractor.

This example demonstrates two important principles:

1. If the operator's task is easy and/or the display is

uncluttered, color provides no performance benefits.

2. If the task is difficult, color coding must be approp-

riately related to the operator task to have value. If

it is not related, it can degrade performance by serving

as a distractor.

56

I

GOOD

C_

C9

&.....ACHROMATIC--- ,,COLOR AS CUEg-. .-- COLOR AS ORGANIZER

40,,--- --* OLOR WITH NO MEANING

POOR ___

INCRFASE IN TRACKItPr, TASK plrFICULTY 0.

Figure 22. Relative Effects of Task Difficulty onPerformance of Simulated Piloting Tasks

as a Function of Different Methods ofColor Coding

TO MINIMIZE EFFECTS OF IRRELEVANT CODING

* Analyze various ways in which information might be

grouped together for various tasks

"* If possible L- r.olor coding as an aid for:

1 ) The most frequently used, or

2) The most difficult operator task

"* Avoid use of color that serves no definable task function

57

COLOR CODING IN HIGH DENSITY DISPLAYS I* Reduces search time: !

1) If target position is unknown, and

2) Target color is known

* Increases search time:

1) If target color is unknown

Effects of Displayed Symbol Density

Display density refers to the number of symbols on a display. When the

display is unformatted to the extent that target position is unknown, the non-

target symbols serve as distractors. If symbols are similar (e.g., all

alphanumerics) the operator may have to examine each symbol to determine

whether or not it is a target. One study reported search times approxi-

mately equal to one fifth of the total number of symbols. When color rroding,

was added, search times were reduced to one fifth of the number of alt.r-

natives of the target color.

Green, B.F., W.J. McGill and 1I. M. Jenkins. "The Time Required to

Search for Numbers on Large Visual Displays, " Report No. 36. LincolnLaboratory, Massachucette Institute of Technology, August 1953, 15 pp.

58It

The relationship between accuracy of locating targets and display density is

shown in Figure 23 for three viewing times.37

To be effective as a code, the color of the target must be known. Without tthis knowledge, multicolor displays may only distract from performance.

This relationship is shown in Figure 24.4

In Figure 25, a comparison between color coding and several shape codes38

is plotted as a function of symbol density. The fewest counting errors

occurred using color coding. The relative superiority of color coding

becomes more pronoiinced as syrnboI. density increases.

r -

USE COLOR CODING TO REDUCETHE EFFECTS OF HIGH SYMBOL DENSITY

* By presenting functionally related items in the samecolor, or

* By presenting "target" data in a unique, prespecified

colorr --g,.4wrig ih'

3 7 W.. and R.J. Christman, "Relative Influence of Time,

Complexity and Density on Utilization of Coded Large Scale Displays,"RADC-TR-65-235, Rome Air Development Center, Rome, N-Y, AD 622-786, September 1965.

38Wolf, E. and M.J. Zigler, "Some Relationships cf Glare and TargetPerception, " WADC-TR-59-394, USAF Wright Air Development Center,Wright-Patterson Air Force Base, OH, AD 231-279, September 1959.

59

o UNCODED0 COLOR CODED

100

90

80 20 SEC VIEWIUG TIME

70

•- 60 r&. 50

4030 --- 10 SEC VIEWIPJG TIM'.

2020 SEC VIEWING TIME

110 10 SFC VIEWPIG TIME

_ iIi

0 100 200 300 400

NUMBER OF ALPUANUMERIF. CIHARACTERS

Figure 23. Effect of Density and Display Exposure Time on Accuracy

I MONOCHROML

A MULTICOLOR - TARGE, COLOR UNKNtWN

MUll_ ICOLOR - TARGET CticuR KK)WN

.4

15 30 45 609

NUMU,1ER 01 SYMBOLS ('N DISPLAY

Figure 24. Effect of Color Coding as a Function of Display Density j

60 I

80

AIRCRAFT SHAPES

',, 60 GEOMETRIC FORMS

'" MILITARY SYMBOLS

40

C>

S20 COLORS

F-"

0 20 40 60 80 100 120

NUMBER OF DISPLAYED ITEMS

Figure 2 5. Counting Errors as a Function of Display Density,Comparing Color Coding With the Three Shape Codes

Coding of Multiple Displa) Sets

When the operator has not one but many color displays in a set, the principles

discussed earlier apply not only 'o each display, but also to the set of dis-

plays as a whole. The designer should be aware that the displays interact

in their effect on the operator. The entire set should be color coded in a

coordinated, consistent way. That is, the meaning associated with a partic-

ular color on one display should be consistent with its meaning across the

remaining displays in the set. For example. if red is used as a conventional

warning signal on one display, it should have a similar connotation whenever

it is u3ed elsewhere. Couflicting uses of color across displays will lead

to operat:r confusion and possibly to misinterpretation of the meaning con-

veyed by the color displayed.

61*1

Another consideration in designing multiple color coded displays is the use

of color to group rotated information appearing on separate displays. Again

this principle is similar to die one discussed in the context of a single display. i

Application of color to a multiple display situation may become clearer if

one views the set as one large display of moderate to high symbol density

rather than several smaller individual displays of low symbol density.

21.

I

ii

t I

I.-

(32

, , i i i i i i i i i i i i i i i i i i i i i i i i i

PART II

APPLICATION OF COLOR PRINCIPLESTO FIGHTER/ATTACK AIRCRAFT DISPLAYS

Prepared by

J. D. Wolf

"J. H. Sandvig

I.

I

ISECTION IV

PART II INTRODUCTION AND OVERVIEW

The objective of Part I is to extrapolate color coding principles developed

in Part I of this design guide to a particular design/operational setting as

an example of how these general principles may be applied. The application

addressed bere is the color coding of electronic displays in fighter/attack

aircraft. An analysis framework is developed that assumes the poitit of view

of a display system designer contemplating the use of multicolor displays,

either in retrofit to an existing aircraft or as components of a new avionic

system under development. With this point of view, our main concerns

are (1) identification of design/operational factors that may influence color

coding, and (2) definition of one or more rucommended coding schemes for

subsequent use in evaluating display media and computatioual requirements.

Background inforatonn -s provided on representative fighter/attack air-

craft displays and their use in realistic mission scenarios. Design and

operational factors judged most relevant to display color coding are identi-

fied. These factors are classified into categories reflecting different

aspects of the system/operational problem typically faced by a designer

during planning and implementation of flight display systems. Issues

associated with these factors and principles and recommendations stated in

Part I are jointly reviewed to define tradeoffs In color application. Based

on these evaluations, color codes recommended for application to electronic

displays in fighter/attack aircraft are presented.

63

|ISECTION V

ELECTRONIC DISPLAYS INFIGHTER /AT'rACK AIRCRAFT

Part II of this report addresses the application of color coding principles

to displays in fighter/attack aircraft. General background information on

representative disjiiays and display concepts is presented in this section.

AIRCRAFT DISPLAY TRENDS

Task analysis and other descriptive data on the following three fighter/

attack aircraft were provided by the Navy in support of this study. These

aircraft are illustrative of the trend toward use of cockpit electronic dis-

plays to replace electromechanical indicators.

"* Vought A-7E: Upgraded version of an aircraft that has been

operational since the late 1960's

"* Northrop/McDonnell Douglas F-18: Advanced aircraft under

development, expected to become operational in the early 1980's

"* VFA-V/STOL: Conceptual design representative of the Navy

Type B V/STOL aircraft expected to become operation in the

1990's

Figure 26 shows the complement and general layout of electronic displays

in each aircraft. Panel displays primarily serving a single function (i. e.,

64

I

i

'I -

A- 7E

a HEAD-UP DISPLAY (HUD)

# THREAT ANALYZER DISPLAY (TAD)RAR

s RADAR DISPLAY 0

F-18 HUD

* HEAD-UP DISPLAY

HORIZONTAL SITUATION DISPLAY (HSD)

' MASTER MONITOR DISPLAY (' WD) r LMIL

M ~ULTIFUNCTION DISPLAY (MFO)

* THREAT ANALYZER DISPLAY

VFA-VSTOL

9 HEAD-UP DISPLA,'

o VERTICAL SITUATION DISPLAY (VSD)

* HORIZONTAL SITUATION DISPLAY VS

* MULTIPURPOSE DISPLAY (MPD)

Figure 26. General Layout of Electronic Displays in ThreeRepresentative Aircraft

65

t

radar and threat analyzer displays) are shown as circles. Head-up displays

and other panel dkplays indicated by squares typically serve multiple

functions or present a greater variety of information.

Electronic displays in earlier aircraft such as the A-7 tend to be single

fuin-tion, b,-t may have various modes tailored to provide information needed

in specifc missiona .hases. For exam'ple, terrain avoidance, ground map,

air-to-ground ranging, and terrain following are common radar display

modes.

From E1'lgure 26 it is apparecnt that the number of electronic multifunction

dh',plays will increase in advanced aircraft. The VFA (vertical fighter

attack) - V/.TOi, instrument panel Is representative of the Navy's advanced

inLegrated display system (AIDS) concept. Although the F-18 panel does

not contain an elect.-onic VSD and nanes assigned to other displays differ

somewhat, this panel implements many of the AIDS features, especially

in lay and aosir-riment of di-play functions.

ADVANC LD INT EGRATED DISPLAY SYSTEM

The AIDS concept will be emphasized here sirce it characterizes the

direction of display system development for advanced Navy aircraft. This

ccncept encompasses a broad range of tec~hnologies including display nmtdia

end processors, data proce.,sing and bussing. and multifunction controls

integrated to provide an ienproved crew/syster., Information interface.

Becaiuse of the continuing development of AIDS for various aircraft and

66

II

mission applications. 35 * 39a singular set of system character-

'5 istics has not yet been defined. However. general features of the AIDS

concept that are relevant to application of display color coding can be

identified. These ar• briefly described in the following paragraphs.

AIDS Display_ ComplementU

STUp to six electronic displays are being considered for integration into the

AIDS systenm,. 40, 42 These are:

e Head-up display (thUD)

* Helnet mounted display (HtMD)

* Vertical situation display (VSD)

* Horizontal situation display (IISD)

* Left and right situation advisory displays (LSAD and RSAD)I

"'"Advanced Display Technoiogy: ANvanced iegrated Moduiar instrumea-tatir,n System (AIMIS " Second Advanced Aircrew Disaym y siun,.Naval Air Test Center. April 191T.

40,"Advanced Integrated Display System (AIDS), " System Design InterimReport No. 3. General Electric Co.. October 1977.

41Dowd, C.A., "F-14 Displays Growt!. for 1980's and Beyond, " Third Aii -

to-Air Fire Control Review. U.S. Air :.'orce Academy, OctobTrT57T

Osterman, L.O. and Mulley, W. G.., 'Anvanced Integrated Displak -vstem(AIDS) for V/STOL Aircraft." A1AA/NAMA Ames v/srOL Co.-Serence.jone 1977.

•'l; '

.4

Although the HMD was not initially considered a component of AIDSP

more recent sources indicate that this display may become an, int- rral part

of the system. 40, 42 The term "situation advisory display" is the name

more recently assigned to the multipurpose displays shown in Figvre'?6.j *

AIDS Concept Description

I

Functions and forms of information presented on the above displays are

summarized in Table 9. The primary anrd secondary functions l1Lted ;,re

a composite of AIDS concepts described in References 35, 39, and 42.

Detailed examples of display information contents and forme-ts can be found

in theRe soure. Fware 2 1., initiay consed a' in ent fI. was audap .

Reference 35 and Illustrates representative formats one all ditsplay

the HMD. Forms of information presentation listed In Table 9 are

for purposes of the present analysis as: d

* Alphanumeric- -letters and numbers

* Symbolic-r-emblematic symbols and line segments

Sensor video- - imagery from sensors such ao racary f nfrcnied, .

and television devices

* Projected ýnap- -map image rear-projected on displa.;

* Direct view of outside wor-ld.--portion of outside .- orld viewed

through HUJD or HMD optical comnbiner

Information forms included on the format examples in Figure 21 are

alphanurceric and symbolic elements on all displays. and sensor video on

the VSD. The following paragraphs further elaborate on general features

of the AIDS concept summarized in Table 9.

• Po~ctci ap-mp iag rar-roeced n isla

0

L0 'o 0 0 0S.-

U' 0

o ol

1 0 c M'

1' 0

~ Vx

CL.

$69

Display Malfunctlons- -Assuming the full complement of six displays listed

in Table 9, each display has at least one backup in the event of a malfunction.

If a malfunction occurs, a second display must serve both its primary and

backup functions by either temporal or spatial sharing of display area. This

could pose a difficult system design problem for applications where both

displays may be needed simultaneously for successful mission performance.

To our knowledge this problem has not been fully resolved.

Other System Malfunctions- -Engine malfunction and advisory information is

presented on the RSAD. 39 The LSAD presents warning, caution, and

recommended course-of-action messages for malfunctions in all aircraft

y.�-t�e-x . Discrete , nessageu appear on the HUI), VSD, and HSD to cue the

pilot that malfunction and corrective-action information is being displayed

on the LSAD. Since the LSAD and RSAD are also used for other purposes.

display-when-needed and automatic prioritization techniques are applied to

present other mission-related information only when required. Examples

are threat detection/classification and countermeasures activation data

displayed on the LSAD and HSD.

Cueing Messages--As with system malfunctions, discretes are presented

on the HUD and VSD to cue the pilot that a critical event (e.g., detected

threat) has occurred, and that related information is being displayed else-

where (e. g.. on the HSD and LSAD).

Mission-Related Modes--Multiple pilot-selectable display moies containing

information and fcrmatting tailored to specific mission phases, weapon

types, or operating conditions are characteristic of advanced electronic

70

J

display systems. Representative HUD formats from Reference 39 are

shown as examples in Figures 27 through 30. Other HUD formats not shown

inlude bombing mode and boresight weapon mode. The following generaliz-

ations can be made with these figures as examples:

* Computer-generated information is a mixture of alphanumerics and

symbology (e. g.. all HUD formats).

* Computer-generated information may be mixed with sensor video

or direct view of outside world (e. g., all HUD formats, but not

shown in figures).

e The same display space may be used for dimensionally different

information presented in a similar form (e. g., altitude and closing-

rate tapes in Figures 28 and 29).

The above comments also apply to the HMD and in most instances to the

VSD and HSD.

Composite Presentations- -- It was previously indicated in Table 9 that as

many as four ATDS displavs may be canable of depiuterng coxn fHter-gn•,-•

alphanumeric and symbolic information, and one or mnore forms of more

complex imagery. Composite or superimposed presentations of various

information forms will be typical. The follov. ing possibilities exemplify

composite display presentations.

SHTJD or HMD (day- good visibility)

-- ComPuter- generated information and direct outside view

H HUD or HMD (niht or limited vsis;ility,) or VSD or B1SD

-- Computor-gene,-ated informption and single or multisensor video

71

-- 2 4 26 28

A I IA I I I

<I

- - 25

250 "

- I .. . .. -2

300 -

XI000BARO<-j

Figure 27. HUD Format, Takeoff/Navigation Mode

20 22 24

r

250 -

-< - -20

300-

IAS X100

Figure 28. iWOD Format, Terrain Following Mode

72

14.I I I I I Ii

-, I

-4

> -.

MAT CFciV I V t L[.I NC.;

I I I I a i I I I RATE

a 1 2 4 MI

Figuire 29. HUD Forilmat, Guided Weapons Mode

,i ja -.

1f'---/-" .- >-

1 -< L

i!, --

Figure 30. I{UD Format, Landing Mode

73

* HSD

-- Computer-generated information and single or multisensor video

-- Cornputer-generated information, sensor video, and projected maps [I .Under day conditions with good visibility, direct view of the outside world

through an HMD or HUD combiner is an inherent component of the informa-

tion contained "or." these displays. Under low visibility or night conditions,

compcr;ex -generated and/or sensor video on a IHUD or HMD produces a

display that appears similar to the same information presented on panel

displ>: s &uch as VSD or HSD. Composite presentations of video from

multiple ser.sors operating in cf'.erent regions of the electromagnetic spec-trui,. a--- "Zkevb in- future sysL,,--s. flaslc requirements - for purpoc- Of

dlisplay are comparable image scaling and accurate superposition of scene

features in the composize image. The HSD offers an additional option of

presenting ;ý projected map image in conjunction with computer-generated

information and sensor video.

Display Mefnia anJ Image Generation

, number of med-i and image-generation characteristics representative of

cux rer.-t -ispiays are anticiu neC to remain as characteristic of future

:nulticcxcr rmea±_ *... - -s -I ne electronic displays in the F-18 and current

AIDS concept are 'atnoce-ray tubes (CI{Ts). The liquid-crystal mediaum

has been recominend&-: fo =:.e AIDS situation-advisory displays. 40 Alp'i-

numeric and symboic i:mag-,s are stroke written en the F-18 CRTs, and

sent, or videt-o is presentet: uSing conventional raster- scan techniqLes. 41

74

With AIDS, there is an increased trend toward use of in-raster alphanumeric40

and symbol generation. These information forms can be generated using

stroke-written, In-raster techniques, or a combination of the two. Figure

30 is an example of a stroke-written display. Figure 31 is an example 43 of

the combined stroke and raster technique as applied to the F-14 VSD. In

Figure 31, stroke writing is used to outline and improve definition of the

in-raster vertical white line and square near display center.

Ambient Conditions of Use

The extreme range of ambient light conditions under which displayed infor-

mation must be legible in fighter/attack aircraft poses one of the most

severe challenges to display designers. Display-surface illumination from

direct sunlight can approach 110,000 lumens/ 2. If direct-viewed back-

ground of a HUD :r HMD is snow illuminated by an overhead sun, luminance

of this background may exceed 34, 000 cd/m 2 .

At the opposite extreme are night operational conditions, where mission

tate detection of targets, obstacles, and landmarks. Maximum display

luminance of less than 3. 5 cd/nr2 is desirable under these condit.ons.

4 3 Elson, 13. M. . "F-14 Uses Digital Display Method," Aviat-ion Week andSpace Technology, July 20. 1970.

*75

-olh le ýtvckk,Gkl vli l

II

I ISECTION VI

DISPLAY USAGE ANALYSIS

Pilot usage of cockpit electronic displays under representative mission

operational conditions is analyzed in this section. Analysis results com-

bined with the more general background information in Section V provide the

basis for evaluating color-application tradeoffs in Section VII.

Five mission segments are analyzed. A taxonomy of electronic display

usage in each mission segment is developed to illustrate the types of pilot

tasKs and v-isual perceptual nnnrntlons- involvped dulring use of these- displays.-_

A display use-frequency analysis quantifies the number of operations per-

formed using the various aircraft electronic displays. Finally, a link

analysis identifies the transitions of visual attention between information

elements on electronic displays and other information sources.

MISSION SEGMENTS

The display usage analysis was completed for mission segments of the three

aircraft introduced in Section V. Task analysis data including pilot-workload

77

44"timelines were provided by the Navy for missions of the A- 7E, the

F-181 45,46 and the VFA-V/STOL. 35 Five mission segments involving

relatively high pilot workload were selected from the following mission

phases:

9 A-7E--Close-air support (CAS) mission; air-to-ground attack phasc

* F-18--Escort mission; ingress and medium-range-intercept (MRI)

phases

* VrFA-V/STOL--Deck-launched intercept (DLI) and CAS mission;

target attack phases

Hi-gh workload segments sunimarizcd below were chosen for the present

analysis because of the relatively frequent ise of a variety of different

displays.

A-7 Ground Attack Segment

This segment begins with weapon delivery preparation including a 10-degree

right turn and a climb from 1500 to 3000 meters. A straight and level

44Klein, T.J., 'Night Display--One-Man Aircraft Compztibility Analysis,

Report No. 2-55900-OR-2806, LTV Aerospace Corp., May 1970.

4 5Asiala, C. F., "F-18 Man/Machine Evalualion Techniques, " Presented at

the American Defense Preparedness Association Avionics Section, AirArmament Division, Air Force Systems Command, Wright-Patterson AFB,Ohio, October 1976.

4 6Asiala; C. F. and Rosenmeyer. C. E., "F- 18 Human Engineering Task

Analysis Report, " Report No. MDC A42 76, McDonnell Aircraft Co..St. Louis, MO. August 1976.

78

interval follows in which the pilot must locate and identify the intended

target and prepare weapons for delivery. A one-half-g nose-over initiates

the delivery dive, during which the identified target is confirmed and the

ordnance is released. The segment is completed by pulling out of the

delivery dive and performing an escape turn.

F-18 Ingress and Intercept Segments

Tile scenario is that of four fighters escorting a small number of attack

aircraft on an interdiction mission. The F-18's prime mission is to protect

the attack aircraft from interception by enemy aircraft. Two mission seg-

ments were chosen: ingress, which is essentially the flight from the forward

edge of the battle area (FEBA); and medium-range intercept (MRI). an air-

to-air intercept of an enemy aircraft using the Sparrow missile.

The F-18 ingress segment of the fighter escort mission involves maintaining

a weave maneuver over the strike aircraft group at constant altitude, noting

the group's position relative to the FEBA and the target, and keeping watch

for enemy threats (i.e., EW and SAMS). The F-18 MRI segment involves

l ........ ....... ........ y in r e ptor. ±radar La-d j im uStcoordinate his attack on the target with his wingman, compute and fly the

I intercept, and launch Sparrow missiles. Finally the pilot must rejoin the

attack group and assume escort responsibilities once again.1VFA Ground and Air Attack Segments1Close-air support includes support of friendly ground operations in close

1 proximity to the enemy. The VFA CAS segment requires locating the

1~ 79

1

- ° w ,_-At

appropriate battle area and communicating with a forward air controller,

then finding the target assigned. The aircraft uses IR sensors and a laser

target designator system, and attacks the ground target once with rockets

and twice with guns. The VFA escapes from the battlefield ax -t after evad-

ing a SAM.

The VFA DLI mission segment includes an air-to-air intercept of a hostile

aircraft threatening to attack friendly forces. The pilot is given information

concerning the location of the target, but retains responsibility for locating

the target on radar and making the final intercept. The pilot flies the com-

mand intercept parameters and uses ECCM techniques. The scenario calls

for the launching of two missiles at the target, probably of the Sparrow type.

This is followed by an attack break and reattack maneuver culminating in

the launch of a Sidewinder missile. The mission segment is terminated upon

the pilot's verification of the enemy aircraft impact.

DISPLAY USAGE TAXONOMY

Taxonomies of electrunic display usage for the above aircraft/mission segment

combinations were defined at the following five levels:

1. Type of aircraft (ie., A-7, F-18, VFA)

2. Segment of a particular mission (e. g., interdiction supersonic

dash, escort medium range intercept, deck-launched intercept)

3. Electronic displays available in the cockpit (e.g., HIUD, HSD, VSD)

4. Pilots tasks required to successfully complete the mission seg-

ment (e. g. . track target, maintain flight conditions, detect hostile

targets)

8I

5. Display elements employed and the pilot's use of display elements

in accomplishing level 4, for example, in accomplishing specific

tasks (e. g., identify target and aircraft symbol, observe discrep-

ancy between the two)

In general, the information necessary to complete levels 1 through 4 of the

taxonomy was provided by a crew tabk-analysis of a crew task time line for

a particular aircraft and mission, Level 5 of the taxonomy was produced by

considering the perceptual requirements of the pilot to accomplish the tasks

classified in Level 4. Terminology used to describe the pilot's perceptual

requirements in Level 5 was taken from a classification of perceptual

processes developed by Berliner et al47 and subsequently revised by48 r

Christensen and Mxlls. These termF;, appiled here to characterize visual

perceptual operations by the pilot, are listed below with definitions devel-

oped for purposes of the present study.

* Detect- - To perceive the presence of some signal not previously

present and not actively sought.

* Inspect-- To peruse; to examine with attention and in detail.

e Observe--To watch; to take note of.

4 7 Berltier, C. D., et al.. "Behaviors, Measures, and Instruments for

Performance E'valuation in Simulated Environments. " Presented atSymposium on Quantification of Human Performance, Albuquerque. NM,

August 1964.

4 8 Christensen, T. M. and Mills, R. G., "What Does the Operator Do In

Comrplex Systems, " hfuman Factors, Vol. 9, No. 4, 1967.

81

4L~1,

* Read--To take in the sense of alphanumerics or symbols.

* Receive--To assimilate passively.

* Scan--To glance from point to point in search of particular

in•jrmation.

SSurvy- -To examine the goneral condition or situation.I

* Discriminate_--To perceive the difference between ele-ments.

"* Identify -- To establish the distinguishing characteristics of

an eiement.

"* Locate--To determine the position of an element.

For example, according to the mutually exclusive definitions above, it was

determined that in the A-7 the functions of HUD for tracking the target would

entail identifying the target and the aircraft symbol and observi,4g the dis-

crepancy between the two. In the F-18. on the other Land, while monitoring

the radar display, the pilot must inspect the display for targets. Thus,

fairly extensive knowledge of the cockpit design and the display formats as

well as the pilot's perceptual needs to accomplish the tasks are required

for this level of analysis.

Results of the display usage taxonomy for the five aircraft/mission segment

combinations are illustrated in Figures 32 through 36.

The large majority of perceptual operations served by elect.-onic displays in

the A-7 CAS weapon delivery segment are associated with the HUD ('3ee

82

CU

02:Ir

C In

0

Onn

83

o11

b.0

84,

ME'IL i.P,[N-.E

S3. -nelay

Figure 34. F-18 Mledium I~arige Intercept--Display

12

,F - • --. - -- M- li

ige intercept- -Display Usage Iaxonorny

85

VA

[OSE AIR

P ,G T SI'L F1_S1 PCE A P,- Hoi S[LfCT"NVSD SL F FOpI ARMAJIE NT

S~" ALM !I P 'MSRTR!Ti'A!,A EN1 S 1

OSRETV C LCT:SCAN. LOCA7E

______T_ ________ OBSIET: OCTE: READ: wA IONý

ART:!;; A RETICI.. S FLSO.LuTE,, 5RFR ARE TRGT Mc ED:OBSER ASI*S CPIRERE CrNAl pýVE:0SRF:0UL PRESS.RE RSVPD:

__________ TRGETAVWEN 7AOGET GTSTATUS

CISCPEP.ANCA IOBSER'J S 1'~ THRUS

R EA D : I A IS C PE P A Cy I E CTLO REN E RL

H IE ADlING

Figure 35. VFA Close Air Support- -Display Usage

L CLOSL A ]P

j~~ mO SIL[CTCN~J CM~[. CHLCK TriPi AT L D, FtL: ART!AY N CCNU[O SI A EC AN) WAPNING CURSOR rT

LOCATE StIN [(ln LCI( SCAN:'~I'RFD: ~ IiAPOMS APSNV ARMMENT FOPIRGOC*

pqAVAIL[BKE U1 IDSPLI IFU Y: (" ANIO URSOR L I CS

.Oa PPESSUPE AREAD: NICM AVAILABLE OBSERVII

_j ITTU FCM UVA ILAW I

Air Support- -Display Usage Taxonomy

87

VFA

DL I

MONITOR FLVTE P[v( INDL ACT ER M0~14 1ct AAOJST A-NALY7F THREAl 1M

A;C STATUS I SCACK I A/ iTAJ$ A ND WARNIV, I L I"'

=ICIMINATF: OBSERVE:IETY SCAN: abAFIF DISCRIMI

O NC symOCa e PI SCREANCY ~ I tIDtR UCP TPREAT I tTHREAT a A/C ARIII. EWE CASS'DL G~~CYRLMSAE[ ESG OMHORIZN SyM~l AS INTENSITY-.. SYMBOLS

08A RVitFA ARTIFICDAhOFZN I j ORION SAIGOBLVI

BT ENACREAD:BEW N

$I ATRSPFEDE

A AL71TUDE* HEADING

Figure 36. Deck Launched Inter(ept--Display U

VIA

XA.IN to ~ PI AL. EP L MrD rCT

.LIYN SMAN I:I I! M tL ALTLPNATLl'- ERCFP If, - Pc,ýATJ Ml £SIL I TT~ ULO

L: LI SCP IM UYT F: PN F' SC A, 7r N, :FY: D IW P IM i ATE:

V i~Li. A,'C SY'13rL 7PFJLSSG0L M* C N D IN-PANGE Cut )L FOP . AVIADC * A

IT AVAAL T1 RAC

LBS[RVE: READ: __________

I B~l WEAN TWO IA NCADABOVE . FADIN

IOCATIOF,

COARSE

tched Intercept- -Display Usage Taxonomy

( 89

Figure 32). The reader is reminded that analysis data for this aircraft

were based on an A-7E equipped with a low-ligit--level television sensor

with sensor video presented on the HUD.

Perceptual operations for the F-18 escort ingress segment (Figure 33) are

fewer and more evenly spread across displays than is the case during the

MRI segment (Figure 34). The F-18 ingress is the only segment analyzed

that does not involve some form of target attack. Although the MRI segment

lasts three times as long as the ingress segment (190 vs. 60 seconds), the

greater number of different piloting tasks during MRI is indicative of higher

pilot workload in this segment. A number of features concerning the use of

-L.j3i "v,o cgm,-l arc e. ~v ion~t fror.jn uiAe d~tlal-ri-is. hi the: nigresssegment, the pilot's primary tasks are to assure that the aircraft is properly

configured to counter an enemy interceptor, to verify passing into hostile

territory, to keep watch for enemy threats of different types, and to main-

tain aircraft control. During the intercept segment the pilot is required to

perform essentially the same tasks, in addition to finding, intercepting, and

destroying hostile aircraft. From Figure 34 it can be seen that these addi-

tional tasks demand that a great deal more information be presentea, particu-

larly on the head-up and multifunction displays, than is necessary in the

ingress segment. The extra information requirements are a direct result

of the additional pilot tasks.

In the ground attack segment of the V-FA-V/STOL CAS mission (Figure 35),

perceptual operations are distributed across the five electronic displays. in

ithe air attack segment, however, a relatively large number of operations

involve use of the RMPD while the VSD is not used at all (see Figure 36).

91

4 '

I

Comparison of similar mission segments across different atrcraft types is

also highly informative. For example, the VFA deck-launzhed intercept

taxonomy in Figure 36 represents a mission similar to the F-18 MRW. The

VFA DLI differs from the F-13 MR1 to the extent that the DLI is purely an

intercept mission, the pilot being given certain information concerning the

target from a command center. The F-18 MRI pilot, meanwhile, is respon-

sfi'e fUr finding targets as well as intercepting and destroying them. Given

this difference, the missions are largely the same once contact has been

made between each aircraft and its target,

In Figures 34 and 36 it can be seen that the RMPD is called upon to provide

informzttAon for a larger proportion of pilots' tasks in the VFA than is the

MFD in the F-18, even though the F-lB's MFD generally serves the same

purposes as the VFA's RMPD. Judging from the more extensive use of the

HUD in the M1RI than in the DLI segment, it is reasonable to conclude that

in the VFA the RMPD is an essential display and that the mission in this

aircraft is flown 'head-down, " while the F-18 intercept mi.-sson relies more

on the HUD and is flown more with vision directed outside the cockpit.

DISPLAY USE FREQUENCY

The second part of this analysis involved determining the frequency with

which each of the visual perceptual operations is employed on each electronic

display and for each of three types of presentations: images of objects

(e.g., objects appearing on FLIR, radar); emblematic symbols (e.g., artificial

horizon, aircraft symbol); or alphanumerics. With data from a task time line

or a crew-task analysis, it was assumed that all the perceptual operations

in Level 5 of the taxonomy (within a task in Level 4) occur simultaneously

92

I1

I!

9

and continuously with that task. Thus if data are available that describe a

pilot's tasks in Level 4, the same data apply to the associated perceptual

operations in Level 5. KResults of this analysis are summarized in Tables 10 through 14. These

tables serve to quantify information schematically illustrated in Figures

32 through 36.

Visual perceptual operations defined earlier appear along the left margin

of each table. The form of information attended to by the pilot is indicated

for each perceptual operation. Abbreviations used are "obj" (objects

appearing in sensor-video imagery or in a direct view of the outside world),

"symb" (computer-generated emblematic symbols and line segments), and

"A/N" (alphanumerics). Following are definitions of other terms used in

Tables 10 through 14:

n1 = Number of pilot tasks involving a particular perceptual operation

(e.g., inspect), information form (e. g., object), and display(e. g., IhUD).

n2 = Number of tasks involving a particular perceptual operation and

display.

n3 = Number of tasks involving a particular display.

N = Total number of tasks involving perceptual operations.

Although rows in these tables are additive, columns are not in many cases

since simultaneous perceptual operations are possible. As an example, in

.Table 10, note the value nl = 77 for both observe/symbology/IIUD and

93

i

TABLE 10. A-7 CLOSE AIR SUPPORT, DISPLAY USE FREQUENCY

°1hrc at

111) Analyzer 1Ojc];Ir IIo\ "P uI

obj 211 (7.7 2 81 (7. 7)

Inspect sym t).l N

n2i- (7.7) 2:i (7. 7)

obj 14 (1.9)) 14 ('1..i)Obscrvc . In h 77 (21.2) 77 (21.2)

AIN

n2 91 (2 -. 1) 91 (25.1)

&•hj

p e'td

:\IN 110 (12 0) 11 ((2.0)

n2 116 (12.0) 116 (312.0)

A IN

n2

obj1 r;ic rimin tt' syntfl) 77 (21.2. 77 2)•I.?)

A/ .'/N

n2 77 (21.2) 77 (21.2)

o0)j 2tt (7.71 2ti (7.7)'" .... •"" .vrlb 19 (4. 1) 1 4 1A/N

n2 41 (11.8) 41 (11,B)

oibjl~ocatu symb ,4 (1.1) 4 (1.1)

./N01)1

r12. 4 (1.1) 4 (1. 1)

obj 42 (ll. I) 42 (11 . i)"1 001d synih 7'il (2 1. 5) 4 (1.1) 112 ( .(i)

A/IN, 116 (32.t) 1 :2 0

94

1,3 2:• (6 .O) 4 ( .1]Z40 ((;;.1 2

9

TABLE 11. F-18 INGRESS, DISPLAY USE FREQUENCY

III l (•Sl) :AIM INl' \ulvyzvu" INow" Slim

A A'N

I IC 2 (1.4 7

(5. ti) 1 (2. !) 2 (5.!l 5 (14.7)

A \N I :: O 1 12. 9) 0 1. m

12 (2,' ) 1 (2.9) 4 (11.a)

,I 2 ( .. ) 2 (1 . 9). ,' ('2. ) 1 (2. 9

1 (2,9) , (I.)

clbj

I njs rI ml i,,li tv -v, l 2 (5'. " ) 2 (1.5 9A /N

i? 2 :.'.) 2 (S. 9)

2 K 9)2 (5. 9)hhI'd (" ,., i (2 .!I) "("..1) 7 (20.6)AIN, 2 (5. 9 1 (22) 2 (17.1;)

A (I'") 1 (2.C) (1) 12 ((2. it

- i ,A4

100(, V..

95

TABLE 12. F-18 MEDIUM RANGE INTERCEPT, DISPLAY USE FREQUENCYI 'lhr , .

1it11) HSI;4) MUD:1 ,IMMl) \n l t 1 .•z:

ob(j 14 (7.') 14 (7., 1-•l•'vm•'lb I,AIN

n2 1-1 (7.,,) 14 (7.!')

ohj 1 (0.6)'' 1 (0.6) 12 (.1)0 c v I. SN-111, 94 (5A, 1) 1 (0..f') -45 (5). 7.. N

n2 95 (5 .7) 2 (1.1) '17 (54. 8

1) t';id zn

A/N 91 (51.4) 1 (0.6) 1 (0.6) 9: (52.-5)

n2 91 (51.4) 1 (0.6) 1 (0.6) 1 (52.'5)

AIN

T12

objD):svriwn~itv[( sý int '94 (!,3. 1) "'4 (53. 1

A- I N

n2 94 (5:3. 1) 914 (1 1. 1

ohj 4 (2.') ) 4 2.1•)Identify In 1) 4 (2-:3) - (2.:)) : (4.5)

A/N

n2 4 (2. 3) 8 (4.5) 12 (;. 0)

l cioct .5'z1)1( 1 (0. 6) 1 (f 6)

12 1 (o, r.) 1 (0.6)

obj 1 (0.6) 2u (11.3) 21 (11.''Total rnh !M (55,4) 5 (2. t) 1 (0.6) 104 (58. 0)

A/ 1 1 (51. 4) 1 (0.6) 1 (0.6) 9:3 (52,5)n: 1 99 (5 5.!1 1 (0.6) 25 (14. 1) 1 (f). 6 1 (0.6) 127 (71 . •)

N 177r) I

1O0(u) (N) •%.

96

TABLE 13. VFA CLOSE AIR SUPPORT, DISPLAY USE FREQUENCY

111'!) VSD P1 1PI) 1%\1 ,D lISIF 1Bo Sum

obj 5 (.3. :3) 5 (3. :1)Insp.ct )symb

n2 5 (3.3) 5 (3. :3)

obj 3 (2.0) 19' (12.5) 22 (14.5)Obser-c symb 15 (9.9) 9 (5.9) 1 (0.7) H (n.3) 33 (21.7)

A IN

n2 15 (9.9) 1 (12.5) 1 (0.7) 8 (5.3) 43 (28. 3)

objR3eAd sviyb

A /N 12 (7.9) 9 (5.9) 14 (9.2) 35 (23. 0)

n2 12 (7.9) 9 (5.9) 14 (9.2) 35 (2:3.0)

obj 3 (2.0) 1 (0.7) 4 (2.6)sC, irb 14 (2.2 14 1.?

A IN

n2 :3 (2.0) 15 (9.9) 1;1 (11.8)

objDisc-irninate symb 12 (7.9) 12 (7.9)

AI/N

n2 12 (7.9) 12 (7.9)

objIdentify symb 3 (2.0) 2 (1.3) 5 (3.3)

A ! N

". 2 3 (2.0) 2 (1. :3) 5 (3.3)

obj 3 (2.0) 14 (0.2) 17 (1 .2)_ ocxate symb 9 (5.') 10 (6.6) 8 (5.3) 27 (17. 8)

A/N

n2 3 (2.0) 14 (5.2) 10 (6.6) B (5.3) :35 (213.0)

obj o3 (2.0) 24 (15.8) 1 (0.7) 28 (18.4)symb 15 (9.9) 9 (5.9) 13 (8. 6) 22 (14.5) 59 (3C .8)

Totpl AiN 12 (7.9) 9 (5.9; 14 (9.2) 35 (23.0)

r. 15 (9.o) 24 (15.8) 9 (5.9) 15 (9.9) 23 (15.1) H6 (56.6)

N 152nl100 (nIIN) a:,

97

TABLE 14. V'FA Dr1CK LAUNCHED INTERCEPT, DISPLAY USE FREQUENCY

I Ii1D "VSD P X\S1I) TAU!'I) 111 )SP 11,(um,

obj 5 (6.1) 5 ((;.)Inspect symb i

.\INn2j5t(6.1) 5 (6.1) 1

obj 1 .3)5 (16.3)Observe svnyb 9 (11.0) 1 (1.2) 13 (15.11) 23 (2 .0)

.A N' N J

n2 9(11.0) 16 (19.,5) 13 (15.9) 3i (46.3)

objeCAd synib

.A /N 9 (11.0) 1 (1.2) 3 (3.7) 8 (9. 6i) 21 (25.6)

n2 9(11.0) 1 (1.2) 3 (3..7) 8 (9. ) 21 (25.C)

objScan symb :1 (3.7) 3 (:. 7)

A/N t6 (9.8) 1 (9. t)

n2 ý8 (9.9) :1 (3.7) 11 (13.4)

obj I

l)iscriminite symb 6 (9.6) 1 (1.2) 1 (9.1) 17 (20.7)A IN

n2 8 (9.6) 1 (1.2) 1(1.9) 17 (20.7)

obj!dentffyv ,vmb 1 (1.2) 2 (2.4) 1 ,'1.2) 4 (4.9)1

A IN

n2 1 (1.2) 2 (2.4) 1 (1.2) 4 (4.9)

objLocate symb 5 (6. 1) 5(6.1)

A/N 1 (1.2) 1 (1.2)

n2 1 (1.2) 5 (6.1) 6 (7.:)

obj 20 (?4.4) 20 (24.4)Total symb 10 (1V.2) :1 (3.7) 4 (4.9) 5 (6;, 1) 22 C(6.8)

A/N j (11.0) 10 (12.2) 1 (3.7) o(9,) 30 (i(;.C)

n:! 10 (12.2) 33 (40.2) 4 (4 9) 113 Q5.9) 60 (73.2)

N 6 }2n!

100(nl IN) lo

9d

discriminate /symbology /IIUD. Seventy-seven tasks involved both observe

and discriminate operations on symbolic informati )n presented on the HUD

display. This is only one instance of simultaneous perceptual operations in

Table 10. Therefore, the value n3 = 236 is not the sum of n2 values in the

HUD display column.

The row sum of n3 represents the total number of tasks involving perceptual

operations on the A-7's electronic displays. The difference N - (row sum

n3) = 363 - 240 - 123 is the number of tasks in the mission segment involv-

ing perceptual operations not directed toward the A-7Vs electronic displays.

All subsequent reference to pilot tasks will be to the N tasks included in this

analysis that involved visual perceptual operatiuns.

Analysis results on the A-7 in Table 10 are summarized below.

* The pilot made use of electronic displays during 100 (240/363) =

66.1 percent of the tasks.

9 The IIUD was used most frequently--on 65 percent of the tasks.

a Perceptual operations on informatton forms were primarily

observe and discriminate symbols (21.2 percent of the tasks)

and read alphanumerics (32 percent of the tasks).

The following summarizes results from the two F-18 mission segments

analyzed (Tables 11 and 12):

"* Electronic displays were used on 35.3 percent of the tasks during

the ingress segment and 71. 8 percent of the tasks during the MRI

segment.

99

"* Most frequently used displays during ingress were MFD (11.8 per-

cent of the tasks), HUD (8. 8 percent cf the tasks), and threat

analyzer (8. 8 percent of the tasks).

"* During MRI the most frequently used Jisplays were HUD (55.9 per-

cent of the tasks) and MFD (14. 1 percent of the tasks).

"• During ingress the most frequent perceptual operations on irnfTrma-

tion forms were observe symbols (14.7 percent of the tasks) and

read alphanumerics (11.8 percent of the tasks).

"* During MRI the most frequent perceptual operations on information

forms were observe symbols (53. 7 percent of the tasks), discrimin-

ate symbols (53. 1 percent of the tasks), and read alphanumerics

(52. 5 percent of the tasks).

"* Each of the F-18's five electronic displays was used at least once

during both the ingress and MRI segments.

Display use frequency data from the VFA-V/STOL analysis xTables 13 and

14) are summarized below.

"• Electronic displays were used on 56. 6 percent of the tasks during the

CAS segment and 73,2 percent of the tasks during the DLI segment.

"• Most frequently used displays during CAS were the VSD (15, 8 percent

of the tasks), HSD (15. 1 percent of the tasks), LMPD (9. 9 percent

of the tasks), and HUD (9. 9 percent of the tasks).

" During the DLI segment, displays most frequently used were the

RMPD (40.2 percent of the tasks), HSD (15.9 percent of the tasks),

and the HUD (12.2 percent of the tasks).

100

* During CAS the most frequent per oeptual operations on information

forms were read alphanumerics (23 percent of the tasks), observe

symbols (21.7 percent of the tasks), locate symbols (17, 3 percent

of the tasks), observe objects (14. 5 percent of the tasks), and locate

objects (11.2 percent of the tasks).

* During DLI the most frequent perceptual operations on Lnformation

forms were observe symbols (28 percent of the tasks), read alpha-

numerics (25. 6 percent of the tasks), discriminate symbols (20. 7

percent of the tasks), observe objects (18.3 percent of the tasks),

and scan alphaawmerics (9. 8 percent of the tasks).

9 With the exception of the VSD during deck-launched intercept, each

electronic display on the 'VFA was used at least four times during

both CAS and DLI segments.

This analysis also highlights other characteristics of the various cockpits,

missions and displays. Results indicate that when Inspect is a perceptual

operation. itis accomplished only on the RMPD or VSD as in the VFA and

F-18 missions, or on the HUD when there is no MPD present, as in the

A-7. Read or observe operations on the other hand, for most missions,

occur on nearly all the displays at some time during the mission. Tables

10 through 14 also indicate that three of the perceptual operations dcLfined

earlier were not applied to characterize pilot visual activities. These are

detect, receive, and survey. The operations receive and survey are appar-

ently defined in terms too general to describe pilot visual activities typically

included in published task analyses. Conversely, the definition of detect

101

was formulated to be more restrictive than traditional definitions of this

term in order to obtain a mutually exclusive category clearly distinct from

inspect, discriminate. and identify. Our definition would apply to detection

of advisory messages, failures, or other contingencies--events of the type

not occurring in the scenarios analyzed.

LINK ANALYSIS

The third part of the display usage analysis was a link analy.-is accomplished

by examining a list of operator tasks for a mission segment and determin-

ing the order in which the displays are referenced in performing the mission

tasks.

A count was made of the number of transitions between any two electronic

displays, or between an electronic display and any "ether" display/control

device, or between a specific electronic display and itself. A link is con-

sidered to be the visual transition from one display to another between tasks.

Of course. it is possible that the same task may have to be !iccomplished

twice in a row, or the same display may be used several times in succcession

4- ,--'"--j,.A• everal ud,-"ren, Ltasks. ,'-iLier case would be an example

of a link between a display and itself. Finally. a proportion of links was

computed between each pair of displays relative to the total number of links

in a mission segment.

Results of this analysis ar• presented in Tables 15 through 17. Link analysis

results are incblded only for the A-7 and VFA mission segments. Data on

the F-18 were in a form that precluded similar analyses on this aircraft.

102

ITABLE 15. A-7 CLOSE AIR SUPPORT, LINK ANALYSIS

YI

To Threat

From , ItUD Analyzer Radar Other

HUD 146' (40.3)*' 2 (0.6) 87 (24.0)

ThreatAnalyzer 2 (0.6) 2 (0.6)

Radar

Oth er 87 (24.0) 2 (0.6) 34 (9.4)

*Number of links**Percentage of total links

(total links = 362)

103

I"

W*,

' I'oTABLE 1(b. V'FA CLOSE AIR SUPPIORT, LINK ANAIYbI,'1

VI' (0. ), t (4.0) "2 (6 [4 , o)

's2 (1 2 ) 14 (,'.2) 1 (0.7) 1 (8.7) t; (4.0)

AI~ *1 1'I 8o 7)3 I~*7

11.V1 4 (2. ,i 6; (4l,t 12 3 t:•t;

S~Number''l of iillks

1'''.•ltl,"O ~ :I irlk: I(rut:li links- = 1;•11

104

ol*11!

104

I

TABLE 17. VFA DECK LAUNCHED INTERCEPT, LINK ANALYSIS

I l'! VSI) R2 M.II'I) LM I'D IISD Other

TFr~on "

0 0 IIUI) __1_____._1 (1.2)

VSI)

13 \111) 1 (1.2) 20(24.7) 1 (1.2) 2 (2.

INIPD 2 (2.5) 1 (1.2) 1 (1.2)

I11S ) I•(9. () '2 (2.5) :1 (:1. 7)

Other b 19. 0) 2 (2.)] 12 (14. )

SNumber of linksI: lerccntage of total links

(total links = 81)

105

Links are defined here as transitions of vis -ial attention occurring between

I

tasks or display information elements. Since the link-analysis summary Itables indicate number and percentage of transitions between various areas

in the pilot's field of view, these data provide basic information on dynamics

of visual activity associated with performance of mission tasks. Links of

the following types are included:

* Between information elements on the same electronic display e

"* Between electronic displays

* Between electronic displays and "other" areas of visual attention

(e.g., from HUD to an electromechanical indicator or switch)

* Betwýýeen arpeas of vrisual aftntinnn not !nrtrnh-ringr e-li-rrnni 1r dnplays

(e.g., from one switch to another)r

In the A-57 CAS mission Table 15 shows that 40.3 percent of the links areo

between information elements on the n3D and that 24 percent are from some

"fother" (nonelectronic display) source to the Il(JD. Combined with the 0. 6

percent of links from threat analyzer to IlUD, these figures suggest that

about 65 percent of the tasks were performed using the HUID. This pro-

portion is consistent with the value n3 =0. 65, shown in Table 10 for the

HJD. Another way of interpreting these figures is to characterize the

percentage values as transition probabilities. If the pilot is accomplishing

a task by reference to the IIUD, the probability that he used the HUD on the

previous task is about 0.40, and the pi 3bability that he will use this display

on the next task is the same. The probbtility that he will next use the HUD,

regardless of the display presently in use, is 0.65.

106

These values for the A-7 CA are in sharp contrast to the corresponding

values for the VFA-CAS mission. For the VFA-CAS. it can be seen from

Table 1S that the probability of th.. next task being accomplished on the HUD

(assuming that the present task employs the HUD) is 0. 007. The probability

that the next task is accomplished on the HUD, regardless of the present

task and display being employed, is only 0.01. It is of course true that the

VFA has several electronic displays to employ in presenting the same

information that the A-7 must present on the ITUD.

Links between information elements on the same display tend to be concen-

trated on the VSD during the VFA CAS segment and on the RMPD during the

DLI segment.

Since the A-7 contains only one multipurpose electronic display (the HIUD)

while the VFA - V/STOL contains five, a substantially greater number of

links to other points of attention could be anticipated in the A-7. This is

not the case. at least in the mission segments analyzed here. Links to

other points of attention accouat for 34, 43, and 27 percent of all links

during the A-7 (CAS, VFA CAS. and VFA DLL segments, rcspectively.

CONCLUSIONS OF DISPLAY USAGE/ ANALYSIS

The following conclusions of the display usage analysis are considered most

"relevant to the application of color as a coding dimension on aircraft elec-

tronic displays. Although these conclusions are stated in general terms, it

is emphasized that they are based on analysis of a limited sample of aircraft

types and mission segments.

107

For a given aircraft, the variety and form of information and per-

ceptual operations performed on this information can vary substan-

tially with mission segment (e.g., F-I18; ingress and MRI segments).

e Presentation of sensor imagery can be on either a head-up (combiner)

or panel display. Displays depicting sensor imagery or direct-view

scene are typically the most frequently used electronic displays

(e.g., A-7's HUD; F-18's HUD and MFD; VFA's .UD, VSD, and

RMPD).

Perceptual operations on information forms most common duringpilot tasks involving electronic displays are: observe symlbols;

read alphanumerics; discriminate symbols; observe objects; locate

symbols; locate objects; scan alphanumerics.

e Advisory messares, failures, and other contingencies did not occur

in scenarios analyzed. Had such events been included, the following

perceptual operations on information forms would become important

because of their critical nature: detect alphanumerics, detect symbols.

* In general the various perceptual operations on information forms are

distributed across multifunction displays with no consistent trend

toward particular operations being performed on specific displays.

e Visual transitions between information elements on the saine display

are as common, and in some instances more common, than transitions

between electronic displays.

* Primary displays (those used most frequently) can be either head-up

or- panel-mounted. The display most frequently used is a functioni of

aircraft type, information assigned to a particular display, and infor-

mation requirements imposed by a particular mission.

I MI3

e Although each electronic display is obviously a critical display when

needed and viewed foveally when in use, in the general case a particui-

lar display should also be considered as a device viewed peripherally

most of the time.

The last conclusion listed above is an important qualifier of other conclusions

drawn from the analysis of dikplay usage. Of the 13 electronic displays

included in this analysis of mission segments involving relatively high

perceptual-iEf ormation workload, only two displays were used during more

than half of the pilot's tasks. The A-7 HIUD had maximum use, in 65 per-

cent of the tasks during the CAS mission segment.

Only tasks having visual per-ceptual components were included in the analysis

(e.g., voice communication tasks were excluded). Of the 808 tasks included,

fully 35 percent did not use any of the 13 electronic displays.

109

SECTION VII

EVALUATION OF TRADEOFFS IN

COLOR APPLICATION

Based on background information and analyses in the two preceding sections,

general principles of display color coding developed in Part 1 are applied

in this section to the coding of electronic displays on fighter/attack air-

craft. Design and operational factors relevant to this area of application

are evaluated to identify recommended coding schemes and practical

con.trait o .u ... color codng.r- Specific recommendations are sumrar-

ized in Section VIII.

EVALUATION FRAMEWORK

Consider the designer contemplating use of multicolor displays, either In

retrofit t- an existing avionic system or as components of a new system

under development. A useful evaluation scheme sbould provide "h's d .signr

with a systematic means for answering the following questions:

* What factors should influence decisions concern~ng color

coding of display information ?

* What are the recommended guidelines for application of Jcolor coding associated with these factors? I

Design and operational factors to consider in color coding aircraft displays

were identified from a review of findings in Part I of this report and the jpreceding Sections V and VI. Factors identified are listed below in five

1110

um I l I •

III

categories. 'l'These categories introduct, an increasing number of factors

to consider as the systexu/operational problem addressed by the designer

becomes defined in greater detail.

1. Pr-escribed Requirements

a Applicable standards

2. Natural Environment

J e Ambient lighting

* Vibration

3. Systen Concept

* Display layout

* Color generation capability

* (Jomnbiner or panel display

* Failure backup availability

* Display when needed

-4. Ihuormatidn (Content

P Primary functions

a Secondary functions

5. Mission Operations

e Mix of information forms

1 Perceptual operations required

o D I)isplay use frequency

a * Mission modes

o Pilot workload

implic,itions of the above factors on display color coding are discussed

below. Designer,- should review tbe following information. in the context

IIl

of their specific application, to aid in evaluating relative merits of alternative

color coding schemes. These coding schemes are summarized in Section VIII.

Prescribed Requirements

Applicable Standards -- Military standard MIl.-STD-1472B prescribes color

codes for transilluminated displays such as indicator and legend lights. Al-

though the current standard for electronic flight displays (MIL-STD-884B)

does not prescribe color codes, future standards for electronic displays

are likely to include codes similar to those for indicator and legend lights.

Standardized codes for increasingly higher priority information are yellow,

red, and flashing red. These indicate cautionary, danger or failure, and

emergency conditions, respectively. Other colors including green, white.

and blue are 5pecified for coding less critical information such as in-

tolerance or nominal system status and advisory information. Of these

colors, green is preferable as a nonalerting color to contrast witfh red ano

yellow. White is more easily confused with yellow, and blue produces

poorer legibility of alphanumerics than either green or white.

Compliance with future color codes for aircraft electronic displays will be

mandatory where applicable. Continued use of red ano yellow codes with

their present standardized meanings is most probable. As concluded in

Part I of this design guide, green is the recommended additional color for

other display information in a basic three-color code. Green provides good

legibility of lower-priorily information while maintaining the attention- getting

value of red and yellow.

112

Natural Environment

Ambient Lighting- -Information on electronic displays in fighter/attack air-

craft must be clearly legible, and chromatic contrasts clearly discernible

under the full range of natural illumination conditions. Under low ambient

conditions, acceptable luminance and zhromatic contrasts must be maintained

while minimizing interference with visual dark adaptation.

Recommendations are provided in Part I for luminance contrast ratio,

chromatic contrast (defined in terms of perceptibly different colors), mini-

mum luminance for color perception, character size and resolution, and

stroke width. These recommendations can most readily be applied to con-

ditions of low or moderate ambient lighting.

As ambient illumination approaches levels experienced in direct sunlight,

colox saturation and luminance contrast on electronic displays tend tc

degrade, Extent of degradation varies with display medium as well as

spectral t.ransmittance o& filter media in ambient-to-display and display-to-

eye light paths. In system development programs these effects should be

rn-easure. for t pccJL i L-cli **~andc conidraio to verfOLfA that* t Vna.

and chromatic contrasts do not degrade below acceptable levels.

Under low ambient conditions, use of display colors other than red does not

have an appreciable effect on visual dark adaptation. Advantages of dark

adaptation from red are relatively small and can be easily lost at minimum22

levels of display illumination required for good legibility.

113

L

Vibration- -Visual acuity can be reduced by aircraft vibrations conducted to22

the crew member's eyes. Since adverse effects of vibration are a complex

function of vibration frequency and amplitude as well as seat and restraint

design, the most practical approach is to experimentally evaluate these

effects for the specific vibration environment of a system under development.

The possibility of vibration causing degraded display legibility is greatest if

legibility is already marginal under static viewing conditions due, for exam-

ple, to use of minimal-sized blue characters viewed with high ambient

lighting.

In general, adverse effects of both high ambient lighting and vibration can

be reduced by avoiding use of the color blue on alphanumerics or symbols

whose shape code conveys most of the character's information.

System Concept

Display Layout--All the panel-mounted electronic displays in aircraft and

concepts discussed in Section V are located on the forward instrument

panel within approximately 20 to 25 degrees of nom.inal line of slgut. Lay-

outs of primary flight displays (HUD, VSD, and HSD) and other multifunction

displays (MPD, MFD, SAD, etc.) conform to formal standards for layout of

electromechanical indicators, or informal standards that appear to be -

evolving for advanced concepts such as AIDS. However, design constraints "

on cockpits of future fighter/attack aircraft may dictate substantial changes

in display layout to accommodate smaller front panels and reclined seat r

positions. Examples include relocation of MPDs to lower center console

or side console- -locations that are both nonconventional and more peripheral.

Also, in concepts where displays serve as failure backups to each other,

11114

I

recovery from a failure is equivalent to revising the display (information)

layout. In a display failure mode, primary information normally presented

in a central location may have to be displayed more peripherally.

Red and yellow in the three-color code described above are cues that high-

priority information is being displayed. The fact that peripheral color

sensitivity of the eye is greatest for white, yellow, and blue suggests that

a different scheme may be more suited to coding priority information on

peripherally located displays. Differential coding in this manner is defin-

itely not recommended. Accepted and standardized meanings of red and

yellow dictate use of these colors to code priority information on all displays

regardless of their panel location. Consistency in application of coding

rules is -also a. we,!-acceptcd lhum-an c .ngineerinig p-actlev tu miinimize con-

fusion and response delay under high work load conditions.

Where peripheral detectability of priority color coded messages is in ques-

tion. use of a centrally located master annunciator is recommended to cue

the pilot that priority information is being presented elsewhere. Color of

this indicator should be consistent with thý. highest priority information being

displayed (i. e., red or yellow). Both the central annunciator and associated

display information m ould be coded flashing red in the event of an emergency

condition.

Color Generation Capability--System designers may in some instances be

required to use electronic displays with a mix of different color ge:.eration

capabilities. Certain displays may have only achroxratic, single color, or

two-color capability. This constraint can develop in retrofit situations, or

because of cost, computational, space, or other design limitations.

115

r

* C

Tradeoffs concerning use of achromatic (black and white) displays and

various common colors of monochromatic displays at e summarized below.

"* Achromatic-- good character legibility but white-level

symbols and alphanumerics would reduce attention-

getting value of yellow

"* Monochromatic red or yellow- -not recommended because

of conflict with standardized use of these colors on multi-

color displays

"* Monochromatic green- -recommended since green is predom-

inant color in previously described three-color code

for multi-color displays

"* Monochromatic blue- -good peripheral visibility but

character legibility is poorer than with other colors

Legibility of blue is improved as this color is desaturated toward white.

The extent to which blue can be desaturated for improved legibility without

introducing problems of discriminability between white and yellow is not

known.

Regardless of which colors are chosen for single and two-color displays,

the benefits of having unique colors to denote both caution and warning

(yellow and red) cannot be realized. If use of one or more displays with

limited color capability is required, the most practical solution may be to

selectively assign information requiring yellow and red coding only to dis-

plays that have at least three-color capability.

116

T~

IIIII IIIII I •,4

It

I

Combiner or Panel Display- -Background luminance and color on a panel

display can be controlled to a large extent by the designer. Cn a combiner

display (HUD and HMD). background luminance and color are primarily a

function of the operating environment. The unique feature of a combiner

display is that uncontrolled luminance and chromatic characteristics of the

real-world background become part of the image seen on a HUD or HMD.

One difficulty that may be anticipated with multicolor combiner displays is

maintenance of chromatic contrast between yellow symbology or alphanum-

erics and real-world backgrounds such as snow or sand. Other examples

of minimal chromatic contrast are green and blue display information viewed

against grass-covered areas and clear sky.

High levels of chromatic contrast between many common display colors and

natural back.-Dunds cannot be anticipated under daytime conditions. Care-

ful selection of spectral transmittance and luminance contrast characteris-

tics is therefore necessary to insure good visibility of all information on

multicolor combiner displys viewed against natural backgrounds.

Failure Backup Availability--Advanced display system concepts such as

AIDS have a sufficient number of displays to allow each to have at least one

back-up in the event of a displav failure (see Section V, Table 9). If one

display serves as a backup to another, problems in color coding may arise

from differences in color-generation capability (discussed above) or the

need to combine dissimilar information from two displays onto one. Use of

color to visually group related information was suggested in Part I and has

potential application to backup-disAay coding in a failure mode. This

application would benefit from use of colors in addition to the recommended

117

three-color code to visually separate functionally different types of informa-

tion on a single display (e. g. . flight systems versus stores managementI

information).

Blue is conditionally recommended for this purpose since it provides goodr

contrast with green, yellow, and red in the three-color code. If aiphanum-

erics in the recoded information are of a size for whichi blue coding may

create legibility problems, viable alternatives to blve are desaturated

orange or desaturated green. The latter would provide the least chromatic

contrast with other more-highly saturated green information. However,

color discriminability for thiis application is not considered as critical as

the information priloritizing function served by the green- yellow -red code.

Desaturated orange is less likely to interfere with the attention- getting

value of yellow and red than a more highly saturated orange color.

Additional applications of desaturated orange are discussed below. This

color is brown in aippearance and, in Munsell terms, has a4 hue of 7. 5 YR

and chroma of /4. Chromaticity coorditiates in the CIE system range be-

tween (x a 0. 47. vy 0. 3 7) and x =0 z 3? *y 0. 3'S0 f or Munsell brit-ht-ne3s

values between 2/ and 8/ The color is attainable on a display device

having the capability for additive mixture of red, green, and blue colors at

CLE-standard wavelengths of approximately 700, 546, and 436 unm, respec-

tively. Although this can be approximated on conventional thiree-guin CRT

displays over a range of lumninance levels, similar capabilities of other

display media were not determined in th!s study.

Display When Needed- -The display- when- needed philosophy is becoming

increasingly necessary in advanced systems to avoid excessive display

118

information density or clutter. Certain critical system and mission even~ts

(e. g. , failures and threats) do not occur with sufficient frequency in the

time span of a mission to justify committing display space on a continuous

basis. Wnen these events do occur, ,nd related warning and advisory in-

formation is presented, quick pilot detection and response is required.

Discussions of preceding factors have emphasized the need to apply red and

yellow colors only to the coding of high-priority information, and to avoid

use of other colors that may detract from the attention-getting value of

these priority codes. Use of a centrally-located master annunciator was

recommended to draw attention to appearance of priority information on a

display. Since these concept.s are intended to minimize pilot detection and

response delays, they should be an effecti e means for improving pilot

performance in systems using the display-when-needed philosophy.

Information Content

Prima_ • Functions- -Examples of primary ".nctions of a display are to

present flight control, Aeapon delivery, stores management, or sensor

information. The list of primary functions in Table 9, Section V, indicates

types of mission-related information that may be presented on AIDS displa)

For the most part, pi i.,,ary functions of the same displays in current air-

craft are similar whoýre the respective displays are included. Each display

contains a subset of mission information that may be similar to, or totally

different from, information assigned to -another displaly.

119

Red and yellow priority coding applies to much of the display information

content associated with these primary functions. Assume for purpcses of

discussion that only the basic green-yellow-red code is used on all displays. jUnder nominal mission-operational conditions, all display information

associated with primary functions in Table 9 would be coded green. Fol-

lowing are examples of events that could cause yellow or red information

to appear, either as newly displayed information (display when needed) or

due to change in value of previously displayed information.

e Flight control function

-- Hligh angle of attack

-- Pullup coinznand durin.g t.rra.n fnllonw [nIr

* Weapon delivery

-- Airspeed, dive angle, or acceleration limits

- Pullup command during delivery dive

* Engine status

-- Low oil quantity or pressure

- - uei main or boot pu.vmvp fphire_

System status

-- Abnormal hydraulic pressure

-- Speed brake failure to retract

* Stares management

-- Failure of stores to release

-- Asymmetric load limits

120

ati

" Tactical situation

-- Known enemy air defense locations

" Threat analysis

-- Threat bearing, range, and type

Color illustrations applying yellow and red codes to some of the above

information are presented in the next section. Effectiveness of the use of

yellow and red codes for the abovc types of information is derived from the

same principles that justify use of advisory, caution, and warning lights

on conventional instrument panels. However, on conventional panels,

,fillUiiiaLiuon of a light typically cues the pilot that information is being

presented on an adjacent instrument or indicator. On multicolor electronic

displays thn priority coding is applied directly to the display information

itself.

W Decisions concerning specific information parameters or variables to the

priority coded, and priorlty-code levels to be assigned to specific system/

operational states, must be resolved during system development. For

instance, the numeric values of oil pressure associated with normal (green),

cautionary (yellow), or dangerous (red) states will vary with the engine

used. Asymmetric stores load information may ¾'stIfy priority coding

only under certain conditions (e. g., landing approach with degraded flight

control system or damaged control surfaces). Coding of threat information

could vary between green and flashing red depending on type, bearing, and

range of threat, as well as countermeasures and performance capability of

the fighter/attack weapon system being developed.

121

If a difficult operational situation develops and multiple out-of-tolerance

conditions occur, sudden appearance of numerous yellow and red display

elements could further confuse the situation. The following approaches are -

recommended to alleviate this potential problem:

* Formulation and use of prioritization algorithms that limit

the number of display-element color changes to only the

two or three of highest priority.

* Limited application of yellow and red coding to only caution

or warning messages, and to display elements representing

variables not under continuous pilot control (e. g., engine -

status and threat information).

A potential application in addition to priority coding is the use of color to

distinguish between basic functional groupings such as flight control, weapon

delivery, and navigation information. General application of color for this

purpose Is not recommended. The number of functional groupings is poten-tially very large. Implicit in this fact is the need for a large number of

colors (other than red or yellow) to code different funcHtna.l ,-Oupngn of

information. Most of these groupings also tend to share common informa-

tion forms: namely, alphanumerics and symbols. If, for example, flight

control information on the HUD formai shown earlier in Figure 29 were

coded one color and the weapon delivery information another, colors of

range and range-rates tapes would differ from colors of speed and heading

tapes. Switching to a landing-mode format (see Figure 30) would produce

a change in colors appearing on the display as well as where various colors

appear on the display.

122 1I .1

Although formulation of coding rules related to all functional Eroupings of

information would be difficult to define, limited use of color coding to

distinguish certain functional groupings from all others ic more practical.

Coding applications of this type are considered in the following two sub-

sections.

Secondary Functions--If a display system concept is configured such that

each display has at least one backup with equivalent coloi generation

capability, color coding could aid in separating unrelated functional group-

ings on the backup display.

Assume, for example, that at some point in a mission, stores management

and engine status are presented on LSAD and RSAD, respectively (see

Table 9). If the LSAD fails, the RSAD must serve its primary and secon-

dary functions by presenting both sets of information. Since display failures

are relatively rare events, pilot confusion that may otherwise result from the

"new" RSAD combined format could he reduced by applying different color

codes to the two sets of information. An effective approach would be to

always apply the same color on all displays to denote secondary-function

information so the pilot can easily identify the functional grouping displaced

from a failed display.

Green is thie recommended predominant color for display primary-function

information. Color alternatives recommended for secondary- function

information were described in th, subsection on Failure Backup Availability.

Op'ions include blue (if legibility is not a problem), desaturated green, and

desaturated orange.

123

Mission Operations

Mix of Information Forms- -Information forms displayed in the AIDS con-

cept are listed in Table 9. Under some mission conditions as many as four iinformation forms may be presented simultaneously. For example, during

an enroute mission phase an HSD could reasonably depict a composite of

alphanumerics (range and bearing to next waypoint), symbols (route, way-

points, and compass), sensor video (weather radar returns), and a projected

map.

More typical is the combination of only alphanumerics and symbols. These

two information forms constit the formatted portions of electronic display

presentations. The pilot must uC able to quickly scan an array of dynamic

displays, (see Part I, Figure 21), and selectively attend to specific alpha-

numeric and symbolic elements needed for system management or, control.

This is especially true under high work load conditions such as the weapon

delivery segments analyzed in Section VI. Figures 35 and 36 indicate the

variet- of diqnlav usage during the brief segments of VFA-V/STOL missions

analyzed.

Fewer electronic displays are involved during similar missions segments

with less advanced aircraft (see Figure 32), but comparable perceptual de-

mands can also occur within the confines of a single display. Under high

work load conditions, quick glances at display elements must frequently

suffice to obtain desired information (e. g., deviations from command mach

or closing rate shown in Figure 29). While attending to flight path commands

124

I

i i i I I I I I I I II I

( and 7F symbols in Figure 29), the pilot should be able to peripherally

monitor other command deviations on the same display. Importance of such

perceptual demands is apparent from Table 15, where 40 percent of the

visual attention transitions analyzed during the A-7 segment were found to

be between information elements on the same display.

Assume again the green-yellow-red code as a baseline. Under nominal

system/operational conditions all alphanumerics and symbols would be

green. This may degrade the pilot's ability to selectively attend to infor-

mation elements during brief glances. Limited use of blue is recommended

to perceptually separate related or adjacent symbolic elements that may

otherwise be confused in a monochromatic presentation.

Examples of symbols for which blue coding could be beneficial are the

command symbols (>) located on HUD and VSD formats in Figures 21 and

27 through 30. Primary ii ^ormation content of such srmbols is associated

with their displayed position rather than shape code. For this reason, the

generally reduced legibility produced by blue should not cause a problem in

111mbot dllilar applicatLLulo.

The preceding discussion has addressed information forms that constitute

the formatted portion of an electronic display image. Other information

forms are direct view of outside world, sensor video, arid projected maps.

These forms tend to have higher information density (information content

per unit display area), and are typically superimposed-on formatted infor-

mnation to produce a composite image.

125

1.

Colors in real-world backgrounds viewed through HUD and HMD combiners

are influenced primarily by spectral transmission of coatings applied to

combiner surfaces. With monochromatic displays, coatings are designed

to reflect a narrow ban.d of wavelengths corresponding to the projected image Icolor. This produces maximum projected-image brightness, and has only

minimal affect on natural colors and luminance of outside scenes. Coating idesign becomes a more complex problem with multicolor displays. The

combiner must effectively reflect all colors of the projectcd display image.

Resulting reduction in transmission of similar colors from the outside scene

may distort scene colors and reduce brightness of real-world objects (e. g.,

flares and landing lights). Thus, rombiner spectral transmission and re-

flection are important design variables that must be considered during

development of a multicolor HUD or HMD.

Projected maps and sensor video images are the two remaining high-density

information forms. These forms frequently have symbolic or alphanumeric

information superimposed or adjacently presented in a composite display

image. Applying the basic three-color code to the VSD in Figure 21, for

example, all computer-generated and sensor video information would be

green with various elements separated only by contrasting luminances and

shape codes. The desaturated orange color previously described under

Failure Backup Availability is recommended for coding map and sensor

images to improve separation and visibility of other superimposed informa-

tion. Desaturated green is a possiole alternative but would provide less

chromatic contrast. Blue is not recommended in this application because

perception of high-resolution map and sensor information would be degraded.

126

St

I

Projected maps used with tactical displays may have an established color

code already applied to distinguish between map detail features. If revision

of map color coding is not a design option, this constraint should be consid-

ered in selecting colors for superimposed computer- generated symbology

and alphanumerics.

"Perceptual Operations Required--The display usage analysis in Section VI

indicated that perceptual operations defined as observe, read, discriminate,

locate, and scan were most common in the relatively high-wortdoad mission

segments analyzed. The perceptual operation, detect, would also be a

critical but not frequently occurring visual activity. (Detect is defined in

Sqepti, VT to include perception of informatiun not previously nresent oractively sought; examples are detection of advisory messages, failures,

and other contingency events).

Many of the color-codlhg principles developed in Part I are applicable here

since the basic concern is with perception of colored stimuli. Guidelines

from Part I have been discussed with regard to other design/operational

factors throughout most of this section. Elements of a coding scheme con-

sisting of three basic colors (red, yellow, and green), with optional

additional colors for information grouping and separation (blue, desaturated

orange, and desaturated green) have been recommended for use on aircraft

electronic displays. The following paragraphs review this coding scheme in

the context of perceptual operations commonly required in fighter/attack

aircraft.

127

J A

Two commonly required perceptual operations (observe and read) would be

influenced primarily by the legibility of displayed information. The pre-

dominant color of alphanumerics and symbols in the recommended code is

green. While this color rates better than blue in terms of its effect on

character legibility, other colors including white, yellow, and red yield

better legibility. A compromise is apparent since the predominant color 1representing nominal system/operational states produces somewhat poorer

legibility, but the colors representing priority or critical states (yellow and

red) produce improved legibility. These differences also lend support to

the use of a color such as low-saturation orange (brown) to code sensor

and map information. This color is spectrally most similar to red, yellow,

and white; but because of its low saturation it is not likely to be confused

with yellow or red.

Three other common perceptual operations (scan, discriminate, and locate)

are concerned primarily with distinguishing one display information element

from another. These elements are typically symbols or alphanumeric

messages. The basic three-color code aids in distinguishing between these

elements only when elements represent different levels of priority or

criticality. Under nominal mission-operational conditions, all symbols and

alphanumerics would be green.

Limited use of blue rather than green coding on selected symbolic elements

provides a means for facilitating the three perceptual operations above. In

event of a display failure, scan and locate operations could be facilitated by

coding functional groupings of information on a backup display with blue,

desaturated orange, or desaturated green. Some discriminate operations

128

involve distinction of symbols and alphanumerics superimposed on maps or

sensor video. In these instances, use of desaturated orange to code the

high density information forms provides a chromatic contrast to aid dis-

crimination of superimposed green elements.

The perceptual operation, detect, is the remaining operation discussed in

Section VI that warrants evaluation here. The process of "detecting, " as

defined in this report, can only occur with a preceding change in some

system or operational state (e.g., appearance of a threat warning or oil

pressure warning message).

Detect operations are more likely to involve peripheral vision since irf or-

mation associated with a state chiange may nut appeaLr in ti± vidcIity of

instantaneous line of sight. Since state changes are also likely to be assoc-

iated with priority information, peripheral visual sensitivity to yellow and

red is a relevant consideration. Peripheral perception of color in the

horizontal plane extends to about + 60 degrees for yellow and about + 30

degrees for red. The more limited peripherai sensitivity to red should be

sufficient for peripheral detection of color if a centrally located master

warning or cueing display is used.

For scan operations, peripheral limits of color perception are of less

relevance. Under the reasonable assumption that a pilot knows which dis-

play to observe in order to obtain the information desired, total subtended

angle of the area to be searched would be less than 20 degrees for an 18 cm

diagonal display viewed at a distance of 50 cm. Data in Part I indicate that

chromatic contrast contributes to substantially reduced search time on tasks

involving location of targets in unformatted images. Improved performance

129

b

Ican therefore be expected in scan operations requiring location of objects

in sensor video or map images if the desired target object(s) can be color

coded. if pattern recognition or multiple-sensor techniques are available

for automatic classification of potential targets in sensor video, targets 1should be highlighted by a contrasting color to facilitate their location.

Similar coding of selected objects on map presentations would be possible jin a system having a computer-generated rather than projected map capa-

bility.

Yellow or red coding should be applied as contrasting colors if objects are

classified as priority information (e.g., known or suspected threats).

Coding of non-priority objects can be accomplished with other colors in-

cluded in the recommended coding scheme. If symboI6 are drawn arouna

non-priority objects to denote their presence, blue symbols would be adequate

on unformatted images of either green or desaturated orange color. If

areas of uniform color are placed over non-priority objects, use of green

high1 ghting is recommended on desaturated orange images and desaturated

orange highlighting on green images.

Display Use Frequency--Analyses in Section VI indicated that display use

frequency varies with aircraft, information assigned to displays, and mission

information requirements. A substantial percentage of visual tasks analyzed

(35 percent) did not involve use of any electronic display.

Generally, display use frequency should not be a primary consideration in

color coding. Recommendations made previously under Display Layout

also apply here. Application of different codes, depending on display use

130

-!

II

frequency, would violate a fundamental requirement for consistency in

application of a selected coding scheme to all displays.

Mission analyses for a system under development may indicate that one or

more electronic displays are used only infrequently during certain missions.

If this is the case, a centrally located master annunciator would become

increasingly important to cue the pilot to priority information appearing on

these displays0

Mission Modes--In addition to their multi-function capabilities, advanced

aircraft electronic displays have multiple modes with information content

and format tailored to meet varying mission-phase information require-

ments. Example of HFTIJD fornrnts for various modes were n.-..Iil sdy

illustrated in Figures 27 through 30. The number of modes and extent of

format change between modes may vary with aircraft, mission, and generic

display type (e.g., HUD vs. SAD).

The recommendation for consistency in applying a selected color coding

scheme to all displays also applies to all modes of a given display. During

the course of system development, a designer can initially select a coding

scheme based on factors discussed earlier in this section. As system

definition progresses and specific display formats for various mission

phases are defined, each format must be colored according to the selected

coding rules. The selected coding scheme should be frequently evaluated

during this process to determine if revisions in coding rules are required

to gain increased benefit from color as a coding dimension.

131

a

Assume, for example, a case where the deFigner initially selects a three-

color code (green, yellow, and red) as sufficient based on preliminary Idefinition of display formats. More detailed information analysis subsequent-

ly indicates that additional command symbols will be required oil a HUD Iweapon delivery format and missile launch envelope symbology will have to

be over]ayed on VSD sensor imagery. These format changes suggest that

consideration be given to use of additional display colors. Possible options

are 1) blue coding of all command symbols on the HUD to perceptually

separate these symbols from other non-command elements, and 2) desat-

urated orange coding of VSD sensor imagery to provide a contrasting color

background for overlayed symbology. If the designer elects to use these

colors for the purposes irndicated, the revised five-color code should be

applied to similar information on all formats.

Predominant colors appearing on a display may change during an operational

mission with progression from one mission phase to the next, or in response

to system/environmental c9ritingency events. In the above example, blue

and desaturated orange colors would not appear unless information of the

type associated with these colors is displayed. Yellow and red colorsapplied as recommended in this report would not appear unless priority

information justifying use of these colors is presented.

Pilot Workload--The display usage analyses in Section VI encompassed tact-

ical mission segments imposing relatively high workload demands on the

pilot. Under conditions of high workload, maximum efficiency in transfer

of display information to the pilot is essential. Quick detection of threats,

critical system failures, and other types of priority Lriformation are

essential regardless of pilot workload level.

132

Color coding schemes developed in this section represent just one approach

to potentially improving speed and efficiency of display information transier.

Other more traditional techniques include use of indicator lights, buzzers,

and flash coding as well as careful attention to display layout, information

assignment, and formatting. Display color coding should be viewed as a

supplement to these other approaches for reducing perceptual workload or

improving pilot-system pcrformance.

All the above techniques including display color coding are of greatest

potential benefit in helping to achieve prescribed system performance

capabilities under peak workload conditions and in event of system degrada-

tion or failure. These conditions and operating modes, as defined for a

particular system under development, should therefoie be the primary

basis for deciding which color coding scheme will be most useful.

Thi' basic three-color code (green. yellow, and red) may be sufficient in

syster, s where pilot workload is not excessive. If peak workload levels are

high, use of additional color codes described earlier under Mix of hIforma-

tion Forms is recommended.

i33

SECTION VTII

SUMMARY OF CODING RECOMMENDATIONS

WITH SAMPLE APPLICATIONS

The objective of Part II of this design guide has been to extrapolate princi-

ples developed in Part I to color coding of electronic displays in fighter/

attack airzraft. Design and operational fa tors relevant to this nrea of

application were evaluated in the preceding section to identify recommended

coding schemes and practical constraints on the use of color coding. The

purpose of this section is to summarize these recommendations. Examples

of coding schemes applied t1) repres'i'• tive ISPuy tH ovd•.

RECOMMENDFED DISPLAY COLOR CODES

Recommended display colors atnd their associated functions arc summarized

as follows:

"* Greeli--The recommended Dredominant color for all display

information not assigned one of the other codes listed below.

"* yallow--Rlez.onmmended for moderate priority or cautionary

information, consistent with standardized us:age of this color.

"* Bled--Recommendud for higher priority information repres-

enting danger or' threat conditions, also consistent with

st-idardized usage of this color. (A third level of information

priority coding can be obtained through use of flashing red for

eir: ,ency o01 highest priority information,

134

I

Blue- -Recommended as an alternative to green on selected

symbols to perceptually separate related or adjacent symbolic

elements that may otherwise be confused. Since blue contributes

to reduced legibility, blue coding should generally be limited

to symbols whose primary information content is derived from

symbol position rather than shape code (e.g., command or

tracking symbols ra'her than alphanumerics).

0 Desaturated Orange- -Recommended as an alternative to green

for coding of sensor imagery or computer-generated map

informaticn. This color provides a chromatic contrast with

overlayed or adjacent symbology. An optional use of desat-

urated orange is to code information transferred from a failed

UU••LL.Li -L, 1. %A J.L • I I -11•:L.UZI Ir14u imfur-

mation normally on the backup display.

0 Desaturated Green- -Recommended as an alternative to

desaturated orange for purposes indicated above. Provides

less contrast with more highly saturated green elements,

but generally improved contrast with yellow and red elements.

ALTERNATIVE CODING SCHEMES

Two basic alternatives arc !-euommended, with various options available for

the second alternative.

The least complex is a three-color coding scheme consisting of green,

yellow, and red. These colors provide the minimum number necessary to

code display elements according to information priority. This coding scheme

135

I

may be sufficient if peak information workloads imposed on the pilot are

not excessive. It may also be the most practical alternative in cases where

design constraints such as media or computational limitations must influence

code selection. I

The second basic alternative is a more complex coding scheme, consisting 1of the three preceding color codes plus one or more of the additional colors

recommended above. Number and functional assignment of these supple- Irmentary color codes can be selected by the designer to meet specific system

requirements and constraints. Use of these codes offers a means for Ipotentially improving efficiency of information transfer to the pilot under

high workload conditions. b

COF D IST "AIN A I

Applicable Standards I

Although current standhrds for aircraft electronic displays do not specify

color code requirements, standardization can be anticipated in the future

as color display media become a more viable design option. Compliance

with applicable standards will be mandatory if imposed by design specifica-

tion. Functions of yellow and red codes recommended abuve are consistent

wh6, current standardized usage of these colors on indicator and legend

lights. As a minimum, extension of current standards for use of yellow

and red (and flashing red) on electro. ic displays can be anticipated.

136

ON

Consistency in Color Code Application

Consistency in application of a selected color coding scheme to all electronic

displays is a fundamental requirement to minimize pilot confusion and re-

sponse delay under high workload conditions. If a designer identifies am-

biguities or other limitations In assignment of codes selected, the coding

scheme should be revised to produce a set of coding rules which carn "2

applied to all displays and display information formats.

'Projected Maps

Projected maps may have an cstablished color code. If revision of map

coding is not a design option, this constraint should be considered in

selecting colors for superimposed computer-generated symbols and alpha-

numerics.

SAMPLE COLOR CODE APPLICATIONS

Electronic Displays

F c 37 to 41 c0t, atcLL exL 'amnples Of Color covdiug schexiies applied to

representative aircraft electronic display formats. Illustrations are based

on art work prepared for this report rather than photographs of actual

displays. Although display element brightness and chromatic characteris-

tics are not accurately reproduced by this process, color code assignments

are apparent in the illustrations.

137

Vigu c :7. .--7HUD ilI

I - . - - --

I -. - - - P,.,-- �

-9

I

Figure 37 depicts information representative of an A-7 HUD navigation and

terrain-following mode. The background image is terrain seen through the

transparent HUD combiner under daytime conditions. The recommended

three-color coding scheme is applied in the top illustration of Figure 37.

In the absence of the yellow "warning indicator" (denoting some cautionary

level of system status) and the red "pullup command, " all display informa-

tion under nominal operating conditions would be monochromatic green.

Flash coding used on the A-7 pullup command symbol could be maintained

for increased attention-getting value. The warning indicator shown in

yellow may also be coded red to communicate more critical levels of air-

craft systems status. The bottom illustrations in Figure 37 is shown in

monochromatic green for comparison with the color coded version.

V,-rscntativ• F1-•160 stores ixnariapement display formats arc shown in

Figure 38. These illustrations are intended to depict night viewing conditions

with only display information, push buttons, and controls around the display

periphery illuminated. Formats depicting information before and after

stores selection are illustrated at the top and bottom of the figure, respective-

ly. Since the display elements constitute routine or non-priority informatica,

all display information would be coded green if the three-color coding scheme

were to be applied. The coding scheme shown in Figure 38 includes use of

blue on all symbolic information to provide a chromatic contrast with green

alphanumeric information.

Figures 39, 40, and 41 illustrate coding variations applied to a complement

of five electronic displays representative of the AIDS concept. Yellow coded

information on the LMPD indicates that a cautionary level of hydraulic

146

pressure exists. Higher priority information is also included in these

examples. Oil temperature on one engine has reached a warning level

(RMPD display) and a threat vector is being presented on the HSD. Desat-

urated green and orange colors are applied to sensor imagery in Figures

39 and 40, respectively. Selected status and command elements in Figure

41 are coded blue to illustrate representative approaches to use of this

color for perceptual separation of related or adjacent symbols.

Electromechanical Displays

Part II of this report has concentrated on color coding of electronic displays.

* It is interesting to postulate how color codes recommended from the present

study might be extrapolated for use on conventional electromechanical

*instruments. Figure 42 illustrates an approach using the green-yellow-red

coding scheme applied to a typical engine instrument.

Green bands indicating normal operating range are common on engine

instruments but not on flight instruments. These bands would be maintained

"* where applicable, as in the example shown. Pointer and scale markings are

"coded green to maintain consistency use of this color on electronic

displays. Marginal or cautionary levels of the displayed parameter are

J indicated by onset of an integral yellow light. u,,t of an integral red

light indicates that minimum or maximum values have been exceeded. Use

of other recommended codes such as blue and desaturated orange would be

of most potential benefit to provide chromatic contrast between elements on

"more complex electromechanical displays, such as attitude-director and

"horizontal- situation indicators.

149

-Ai1

Advanced aircraft designs may include both electromechanical instruments

and multicolor electronic displays. If so, consideration should be given to

the feasibility of implementing a selected color code on all visual information

sources to achieve an overall consistency of functional color usage in the

cockpit.

152

I

REFERENCES AND BIBLIOGRAPHY

lI "E 1ý EN C, iS

1. Haeusing, M., "Color Coding cf Information on Electronic Displays,'Proceedings ot the Sixth Congress of the International ErgonomicsAssociation, 1976, pp. 210-217.

2. Bishop, 11.1P. and M.N. Crook, "Absolute Identification of Color forTargets Presented Against White and Colrcd Backgrounds," ReportNo. WADD TR 60-61l, Wright-Patterson AFB, ON, March 1961.

3. Smith, S. L., ''Color Coding and Visual Separability In InformationDisplays," Journal of Applied Psyc'hology, 47, 1963, pp. 358-364,

4. M-yers, W.S., "Accommodation Effects In Multicolor Displays,"Report No. AFFDL TR-67-151, Air Force Flight Dynamics Laboratory,Wright-Patterson Air Force Base, 011, AD 826-134, December 1967.

5. Ellis, B., G.J. Burrell, .. T. Wharf and T. D. F. larokins, "TheFormat and Colour of Small Matrix Displays for Use In 1tigh AmbientIllumination, " Royal Aircraft Establishmeini Technical Report 75048,March 1975

6. Ericksen, C.W., "Multidimensional Stimulus Differences and Accuracyof Discrimination, " WADC 'I'll 54-165, Wright Air Development Center,W,-ight-Patterson Air ,'orce Base, 011, 1954.

Q1 C I. It' CV \ A F ),b ~~.1f4n'.r ~,iT .t'c~oFl~l

In the l~egrility of Televised Displays," Report No. ESD-Tl-66-299,The MIITIRE Corporation, Al) 640-571, September 1966.

8. McLean, M:.V., "Brightness Contrast, Color Contrast, and Legibility,"HumanTI'actors, 7, December 1965, pp. 521-526.

9. llegli', J., NAVSIIIPS 1Display Illumination Design Gud_ Section 11:liunian Factors. Naval Electronics Laboratory Center: San Diego, (CA,1Y97:3.

153

REI'PE IENCE'S (continued)

I •10. Semple, C.A., JIr. 0 R.. . Ileapy, I..J. Conway, Jr. and K.T ' lV1rnette,

"Analysis of Human Factors Data for Electronic Flight Display Systems,

Technical Report No. AFDI)--TH-70-174, April 1971, 570 pp.

11. Pollack, J. D., "Reaction Time to Different Wavelengths at VariousLuminances, " Perception and Psychophysics, 3, 1968, pp. 17-24. I

12. Tyte, 1. , J. Wharf and 13. Ellis, "Visual Response Tinmcs hi I igh

Ambient Illumination, Society of Information DisplaXs Digest, 1975,pp. 98-99.

13. Htaines, R.\4., L..l, DawsoaI T. Ga.van and L,.. l. Reid, "'RSptoniSe

Time to Colored Stimuli In the Full Visual Field, " h•eport No. NASATN D-7927, NASA Ames Research Center, Mloffett Field, CA, March1975, p. 27.

14. Connollyo 1).V., G. SIanli'lr and F. Champion, Color I")isplay "Jv'alu -

ation for Air Traffic Control, " I'eport No. FAA-RI)-75-39, Federal IAviation Administration, Washbigton, D)C, May 1975, 41 pp.

!5. Hanes, l.M. and M.V. lihoad(S, "Color 1dentitication as a I Function ofir ~ Po.ietynd PA \' lihod49, Colo

E'xtended Practice, " Jou-nal of thIe Optic.l SOCiet'Vof Aeic i, 4.9, 1959,pp. 1060-1064.,.

16. Shontz, W.I)., G.A. Trumm and I..G. Williams, "Color Coding for

Information Location," '' I umaxI Factors, 13, 1971. m). 237- 246.

17. Teichner, W.T. and Erebs, .. J., "L1aws of Visual Choice Bu actioný

"Time, " Psychological Review, 81, 1974, pp. 75-98.

1 H1. 1itt, W.I)., ''Ani Evaluation of Five Diffverent Abstract (,1,)dii;, PdethIods,

Hlumnan ,actors, 3, 196J, pp. 120-130.

19. Baker, C. A. and W. F. Grether, "Visual Presentat ion of infornmat ion,Wrigiht Air DIevelopment Center, WAI)G.-TRI-541160, AI)13-064, '-904. "

20. Cook, T.C., '"Color Coding--A litview of tilC 1iteat 'r, '' I. Armn'vH luman EIngineering laboratorv., 1111'l, Tch Note 9-74, Al: ,•' id l'royilig

(Ground, NI), November" 1974.

1-54

I

REFERENCES (continued)

21. Wald, G., "Blue Blindness In the Normal Fovea,' Journal of the OpticalSociety of America, 577, 1967, pp. 1289-1303.

22. Meister, D. and D.J. Sullivan, "Guide to Human Engineering Designfor Visual Displays," Contract No. N00014-68-C-0278, EngineeringPsychology Branch, Office of Naval Research, Washington, DC,AD 693-237, 1969.

23. Rizy, E. r., ':Dichroic Filter Specification for Color Additive Displays:II. Further Exploration of Tolerance Areas and the Influence of OtherDisplay Varicnbles," Report No. RADC-Tli-67-513, USAF lhome AirDevelopmenm Center, AD 659-346, September 19S7.

24. llilgcndorf, R. L., "Colors for Markers and Signals: Inflight Validation,AMRL TR-71-77, Wright-Patterson Air Force Base, 011, AD 737-901,197 1.

25. Airerew Station Signals Military Standard MIL-STD-411D, 30 June1974 with Notice 1, 30 August 1974. Department of Defense: Washington,DC.

26. Colors, Aeronautical Lights and Lighting Equipm nteneral Require-ments for, Military Specification MI11-C-25050A (ASG), 2 December1963, with Amendment 1, 30 September 1971. U.S. GovernmentPrinting Office: Washington, DC.

27. Christ, R.E. and G.M. Corso. "Color Research for Visual Displays,Technical Report No. ONII-CR213-102--3, July 1975, 108 pp.

28. Christner, C.A. and II.W. Ray, 'An Evaluation of the Effect of SelectedCombinations of Target and Background Cobing on Map Reading Perfor-mance--Experiment V," l]uman Fastors, 3, 1961, pp. 131-146.

29. Alliusi, 12.A. and P.Y. Muller, Jr., "Verbal and Motor Responses toSeven Symbolic Visual Codes: A Study In S -R Compatibility," Journalof Experie-.entvl Psvcholog_, 55, 1958, pp. 247 -254.

155

REFERENCES (continued)

30. Garner, W.R. and C.G. Creel-man, "Effect of Redundancy and Durationon Absolute Judgments of Visual Stimuli, " Journal of ExperimentalPsychology, 7 1964, pp. 168-172. I.

31. Ericksen, C.W. and H.W. fake, "Mlultidim'cisional Stimulus Differencesand Accuracy of Discrimination, " Journal of Experimental Psychology,50, 1955, pp. 153-60.

32. Saenz, N.E. and C.V. Riche, Jr., "Shape ano Color as Dimensions ofa Visual Redundw'it Code," Human Factors1 16, 1974, pp. 308-313.

33. Kanarick, A.F. and P.C. Petersen, "Redundant Color Coding andKeeping-Track Performance," Human Factors, 13, 1971, pp. 183-188.

34. Green, B.F. and L.K. Anderson, "Color Coding In a Visual SearchTask," Journal of Experimeniita_ y .. i., 5i( lu), 195b, pp. 19-24.

35. Linton, P. M., "\' FA-V/STOI. Crcw Loading Analysis, " Report No.NADC-75209-40, Crew Systems Department, Naval Air DevelopmentCenter, W•arminster, P;A, 15 May 197b.

36. OG-'en, B.F., W. '. McGill and I], Y. Jenki•s, ''The Time Required toSearch for Numbers on Large Visual Displays," Report No. 36, LincolnLaboratory, Massachusetts instituie of Technology, August 1953, 15 pp.

37, Dyer, W.i 13 arid R.J. CIiixilsntIaI, -'Relatie Influence of Time, Cornplexity and Density on Utilization of Coded Large Scale iDisplays,RADC-TR-65 -235, Rome Air Development Cen er, Rome, Al) 622-786, September 1965.

38. Wolf, E. and MI.J. Zigler, "Some RelationshJips of Gla-C alnd TargetPerception," \VAI)C-TIR-59-394, USAITF Wright Air 1)cvulopnient Center,

Wrighdt-Patterson Air Force B]ase, 011, AD 231-29719, Scptenb1 1959.

39. "Advanced Display Technology: Advanced Integrated Modul.ar hitItrufllen-

tation Sys Lem (AIMIS), ' S-cond Advanced Aircrew DiSLJ)ioi•O1i ur, §Naval Air' Test Center, April 1975. I

1 56 I

REFEI"l'REI'N CE'.S (con1tinueCd)

40. "Advanced Integrated Display System (AIDS)," System Design InterimReport No, 3, G(eneral Electric "'o., October 1977.

41. D)owd, C.A., "F-14 Displays Growth for 19 80's and Beyond, '' ThirdAir-to-Air b'ire Control Review, U.S. Air P"orce Academy, Octobev 1977.

42. Osterman, L.C. and W.G. Mulley, "Advanced hitegrated Disp]laySystem (AIDS) for V / STO1, Aircraft, " A1AA/ NASA-Aiiies V/ S'r1,flConference. June 1977.

43. Elson, B. M., "'F-14 Uses Digital Display Method, " Aviation Week atndSpace 'Technology, July 20, 1970.

44. Klein, J'.J., "Night Display--One-Man Aircraft Compatibility Analysis,Report No. 2-55900-01-2806, 1,TV Aerospace Corp., May 1970.

45. Asiala, C. F., ''1'-18 Man!Machine Evaluation Techniques, " presentedat the Anieri-an Defenbe tlreparedness Association Avionics Section,Air Armament D)ivision, Air Force Systems Comialid, Wright- PattersonAt'B, Ohio, Octobe-' 1976.

46. Asiala, C.FI'. and tosenniuyer, C.E,. , "F-18 1lunian Engincering T''skAnalysis Report, " Report No. MI)C A4276, McDomnell Aircraft Co.,

SSt. Louis, MO, August 1976.

47. Berliner, C. D.s, et al., "eihaviors, Me 50it r-s, anud h2 m en, t s forPerformance I.'valuanion In Simulated Entvironments, " preenteod atSylitposiumIl On Quantification of Itma''I lPertOrllance, Albuquc qut-, NNi,August 1964.

48. Christensen, J. M. and 11. G. Mills, "Vhat )Does the Operator D)o InComplex Systems," I Iliu-n Factors, Vol. 9, No. 4, 1967.

157

ii'1111ý I-TO R APH

1. "Advanced Display TeceLnology: Advanxe Integrated Modular Instrumen-tation System (AIMiS), " Second Advaan,,'eo Airceew Display Symposium,

Naval Air 'rest Center, April 1975,

2. "Advanced Integrated Display System (AIDS), ' System Design InterimRepor-t No. 3, General Electric Co., ()ctober 1977.

3. Advisory Group for Aero:space Research and Development, "AircraftInstrument and Cockpit Lighting by Red or W\hite Light, " AGARDConference Proceedings. No. 26, October, 1967.

4. Advisory Group for Aero.-pacc Research and Development, "ColorVision Requirements In DWffcrent Operational Holes, " AGARD ConferenceProceedings, No, 99, Mday 191''2.

5. Allport, D.A., '.'Pariillel Encoding Within and Ietween ElementaryStimulus Dimensions," Poerc -tion andlPsycho.h.psics, 10, 1971, pp.104-108.

6. Alliusi., E.A. and P.IF, Muller, Jr., "Verbal and Motor Responses toSeven Symnbolic Visual' Codes:: A Study In S-11 Compatihility, " Jour'nalof lxperimental Psvcho__gy, 55, 1958, pp. 247-254.

7. Anderson, N.S. and P.M. 1itis, ''Airount of hlfoat'itatin Gained DlUringBrief Exposures of Numerals and Colors. " Journnl of of',pnirn... .1Psycology_, 56¢, 1.95,a pp. 362-369..

8. Aziala, C.?F., 'F.-18 rxa/l/lachinc l:v\-luation Teliniqu, '' presentedat thit American Defense ]'ri:parediwess Association Avionics Seuclion,Air Arnmamenit Division, Air lorcW :S3"tcns Command, Wright-Patterson AtV,. 011, October 1t9(7.

9. As:a'.-), C.FV. and Iosevne'er0. "Y'., '"'-18 1Human E]nginLeering lpaSkAnalysis Relport, '' Relport No. lNiI)(! A4276, l eIDoumell Aircraft ('o. ,

St. Louijs, M0 August 1976.

1511

111 I O,1It A1PHY ly (contilUed)

10. Baker. C. A. and V. F. Grether, "Visual Presentation of Information,Wright Air Development Center, W ADIG-TUX- 541.60, AD43 -064, 1954.

II. Barker, E'. and M. Krebs,"Color Coding Effects on Human Performance:An Annotated Bibliography," Office of Naval Research Report Nol ONII-CU213-136-1F, April 1977.

12. Barmack, J. E. et all., ,'luinani Factors Problems in Computer G(eneratedGraphic Displays," Institute for Defense A'ial.,sis, S-234, AD 636- 170,April 1966.

13. Barnes, J.. '# 'he Effect of Cockpit Lighting S?'stemns on MulticoloredDisplays, " Report No. 30-70, Hluman Elngiieering Iahoatorks,Aberdeen Proving Ground, Maryland, December, 1970, 33 pp.

14. Bartlett, N.I1., T. G. Sticht, and V. I'. Pease, "Effects of WavelengthGand ltetinai Locus on i he 1 utioj Ti.. to -Uld. ad ff-r...t Stl-1t...,' illJournal of I.xperimenlil ' Psychology, 78, 1968, pp. 699-701.

15. Bedford, H.E.C. and ( W.\',. \V,'szecki, ''\Vav.elngth 1)iscI-inliiUi.iOnl forPoint Source, '" J ournal )f tlhe Optical Societ v of Am erica, 48, 1958.

16. Berliner, C.DI)., tt al., "Beha'viors, Mleasurt's, and Illst lrunL11'Ilts f'of"

Pecrformnance Evaluation ini Simulated E".lnviroolmllelets, " I' escWl ('d zitSynllmosit, a on Qlantit'iCaItionl of IIm1an I'eriforinance, All)ujutrqiuc, N\I,August 1964.

17. lBeyer, 1H. , ''A li'niiltcd Study of thilt Trlade-off Betwe.enl l.umilinalce andColor Coding ill Electronic Aircraft Displays," Guidamce and Contcollnisplays, AGfAR Conference P'roceedii•gs, No. 96, FtbruA rv 1972,

18. Beyer, R., II.D. Schenk nild] E'. Zeitlow, "Investi.igatiotis on the I1I.'v:Id-bility and Interpretability of Electronic l)isplays: Illvet igations on thte.ft'fectiveness of llrightiness Coding and 'olonir Codir.g of l )iV-yE-lements," 'Royal Aircraft Establishment, l ibrarv atioi No,1641, 1971, 80 pp.

19. l~ioaStr'onautis DIata 10ook. W\ashi'igton, IDC: Nationial Acro'imwutics

and Space Adminstration, SP-3006.

159

'.

113113 LIO; f(APtlY (continued) i

20. Bishop, H.P., "Separation Thresholds for Colured Bars With VariedLuminance Contrast," Pychonomic Science, 5, 1966, pp. 231-238. -28

21. Bishop, H.P., "Separation Thresholds for Colored Bars Without ".Luminanco Contrast," Psychonomic Science, 4_, 1966, pp. 223-224. j

22. Bishop, 11.F. and M.N. Crook, "Absolute Identification of Color ftrTargets Presented Against White and Colored Backgrounds," ReportNo. WADD TB FO-611, Wright--Patterson AFB, 011, March 1961.

23. Bzedern, W., "Tri-Color Cathode 3,ay rTube," Rorne Air Developraent ¶Center, Air Force Systems Command, Griffiss Air Force Bast, N-Y,1970.

24. Brooks, 1, , "Search Time and Color Coding, " Psychonomic Science,2, 1965, pp. 281-292.

25. iryden, J.E., "Design Considerations for Computer-Driven C.RT2 Displa'vs, Computer Design, March 1969, pp1. 38-46.

26. Brydt'n, J.E., "Performance of Phosphors, I'sed ini CRlT Displays,"Raytheon Company, Lexington, M\IA, .cpt.etbrvn 1065.

27. Burdick, D.C., L. Mi Chauvctte, J.AM. Dcl, ,,! A. ., Goins, "Color

Cathode Ray Tube Displays In Cc' iat lnfoinutier. cnlenrs," U.S.qNaval Research Laboratory iepo.t Nt,. 634, ')65.

28. Cahill, M. and I.oC. Carter, Jr., "'Color Code Size for Sca.rchingDisplays of Different Density, Hulmnan Fact•ors, 18, 1976, 273-280.

29. Cavunius, c.!{., "''lumaan Visual Acuity 'Measured with ColoriedStimuli," l'Rport No. 1151-1111 -6518-(:n, luman Sciences llReocrcli,!nc., McLean, VA, Al) 472-2'53, Septen-mb,.)r 19C5.

31]. Cavontus, C. P., "-I he 'Effect of Wawvelngth on Visual Acuity," ' eport INo. EHF-I1fl-1/67-Cl,, Eve lesearec Foundation, lPethesda, N1\1),Au) 646-575, .lanuinry 1967.

160

,L

BIBLIOGRAPHY (continued)

31. Cavonius, C.R., R. ltilz, and J.H. Kravitz, "Chromaticity mdLuminance Effects on Visual Detection," Report No. ERF-1R-2/68-CR,Eye Research Foundation of Bethesda, Bethesda, MVID, 1968, 43 pp.

32. Cavonius, C.R. and A.W. Schumacher, "Human Visual AcuityMeasured With Colored Test Objects," Science, 152, 1966, pp. 1276-1277.

33. Chapanis, A. and R.M. Halsey, "Luminance of Equally Bright Colors,"Journal of the Optical Society of America, 45, 1955, pp. 1- 6.

34. C0ase, W.D., "Effect of Color on Pilot Performance and TransferFunctions Using a Full-Spectrum, Calligraphic, Color Display System,A1AA Vision Simulation and Motion Conference, Dayton, OH, April 26-28, 19#'6.

35. Chisim, G.T., "Color Di`scrlmination and Chart Reading Under 'Redand Low Intensity White Lights," Aircraft Instrument and CockpitLighting by Red or White Light, AGARD Conference Proceedings,No. 26, October 1968.

36. Christ, R.E., "Review and Analysis of Color Coding Research forVisual Displays,'! Human Factors, 1ý, 1975, pp. 542-570.

37. Christ, R.E. and G.M. Corso, "Color Research for Visual Displays,"Technical Report No. ONR-CR213-102-3, July 1975, 108 pp.

38. Christensen, J.M. and R.G. Mills, "W/hat Does the OperaLor Do InI Complex Systems," Htuman Factors, Vol. 9, Noe 4, 1967.

39. Christrnan, 1R.J., "Specification of Primary Intensities for Seven-

Color Additive Displays, " Rome Air Developmenvt Center, RADC-TR-68-319, AD 674-589, July 1968.

40. Christner, C.A. and II.W. Ray, ''Ai Evaluation of the Effect ofSelected Combinations of Target and Background Coding in Map ReadingP11crformance--Experiment V," h1uman Factors, 3, 1961, pp. 131-146.

1(31.,

IG!f

BIBLI OGRAPHYl ( con inued)

41. Cohen, J. and V L. Sanders, ''An Experiment on Dial Coding," WADCTechnical Report. 52-209, Wright-Patterson Air Force Base, 0!l,September 1952.

42. Connolly, D.W., G. Spanier and F. Chamnpion, "Color Display Evalu-ation for Air Traffic Ccntrol," Report No. F'AA-RD-75-39, Federal IAviation Adminstration, Washington, DC, May 1975, 41 pp.

43. Connors, M.M. and J.S. Kiniiney, "Relative Red-Green Sensitivity as aFunction of Retinal Position, " Journal of the Optical Society of America,52, 1962, pp. 81-84.

44. Connors, M., "Luminance Requirements for Hlue Perception In SmallTargets, " Journal of the Optical Socicty of America, 58, 1 968, pp.258 -263.

45. Connors, M.M, "J.uniinance Requirements forx ute Perception andIdentification for a Range of Exposuvie Durations, " Journal of the Li

Society of AEturica, 60, 3971), pp. 958-965,

46. Conover, D. W., "The Amount of Information in the Absolute J udgInLnt-tlof A\unsell IHues," Report No. WA1)C 58-262, Wright Air Developiwn)eWCenter, Wright-Pallerson Air Force lO,-:c0 011, AJune 1959, 48 pp.

47, Conover, D. W. and C. L. Kraft, "The Use of Color in Coding I)iaplavs,i(eport No. kAI)C TR 5'-471, Wriglht Ax 1.)evclopment Center, -"

Patterson A"II, OI0, AD 204-214, (X'obtIr V)58.

411. Cook, T. C., "Color Coding- -A Reviev, of the .J .t ertur," ' ."S. Arm.,Iluman Engineering ILaboratory, IIEI. Tech Note 9-74, Aberdecu I 't)V'ilifn

Ground, MI), November 1974,

49. L)avis, J.)A. , "''ecent Advances In Cathode Ray 'I ubc Di.splay ])evi ce7,,"Recent Advances In 1)isplay Media, NASA-SP-1ig, September 1967,pp. 25-39.

50. Dooley, iR.P. and L.E. Hlarkins, "Functional wid Atteition-GettingEffccts of Color on Graphic ComImunicationI," 'PerceLtu:,L and Mloll.Skills, 31, 1970, pp. 851-854.

162 I

BIB3LIOG'iAPHY (continued)

51. Dowd, C.A., 'E-14 Displays Growth for 1980's and B3eyond," ThirdAir-to-Air Fire Control Review, U.S. Air Force Academy, October1977.

52. Dudek, R.A. and G.A. Colton, "Effects of Lighting and BackgroundWith Common Signal Lights on hluman Peripheral Color Vision, " HumanFactors, 12, 1970, pp. 401-407,

53. Dyer, W.R. and R.J. Christnw•n, 'Relative Influence of Time,Complexity and Density on Utilization of Coded Large Scale Displays,"RADC-TR-65-235, Rome Air Development Center, Rome, NY, AD 622-

786, September 1965.

54. Ellis, B., G. J. Burrell, J.oH. Wharf and T.D.F. Harokins, "TheFormat and Colour of Small Matrix Displays for Use in High Ambient

March 1975.

55. Elson, B.M. 'M "F-14 Uses Digital Display Method," Aviation Week andSpace Technology, July 20, 1970.

56. Ericksen, C.W., "Multidimensional Stimulus Differences and Accuracyof Discrimination," WADC TR54-165, Wright Air Development Center,

Wright-Patterson Air Force Base, 011, 1954.

(.'7 t',.4 L.- C IXJ, 0 ,,- ,. TI U! , IT~llrn'. Ulv-h lt'ir'irnp,z nvinnnl gfrn'i,1,,c T-h'f'fnrc-.'nrc

and Accuracy of Discrimination, " Journal of Experimental Psychology,50, 1955, pp. 153-160.

58. Evans, R.M., "Variables o-f Perceived Color," Journal of the OpticalSociety of America, 54, 1964, pp. 1467-1474.

59. Farrell, R.J. and J.M. Booth, "Design l]andbook for Ihagery Inter-

pretation Equipment," Boeing Aerospace Company, December 1975.

60. Colors Aeronautical Lighting, Federal Supply Agency, General ServicesAdministration. Federal Specification No. 3, 21 March 1951, .NithAmendment 1, 27 August 1951.

163

BIB3LIOGRAPHY (continued) I61. Fowler, R.D. and DB. Jones, "Target Acquisition Studies: (1) Trans-

ition from Direct to TV Mediated Viewing: (2) Target Acquisition jPerformance: Color vs. Monochrome TV Displays," Technical ReportNo. NR 196-071, Office of Naval Research, AD 736-244, January 1972,29 pp.

62. French, R.S., "An Investigation of Target Enhancement Through Useof a Multi-Phosphor Cathode Ray Tube, " Engineering PsychologyBranch, Offi 'e of Naval Research, Washington, DC, Al) 652-980,May 1967.

63. Galves, J. and J. Brun, "Colour and Brightness Requirements forCockpit Displays: Proposal to Evalu•te Their Characteristics."Paris: Thomson CSF Electron Tube Group, Lectures No. 6, AGARDAvionics Panel Technical Meetirns, on . Fltni,c Displys, 197- .

64. Garner, W.R., The Processing of Informnation and StructurIe. NewYork: Wiley 1974.

65. Garner, W.R. and C.G. Creelman, "Effect of Redundancy and Durationon Absolute Judgments of Visual Stimuli, " Journal of I.xperimentalPsychology, 67, 1964, pp. 168-172.

66. Geldard, V.A., The Human Senses, Second Edition. New York: Wiley,1972.

67. Green, B.F. and L.K. Anderson, "Color Coding in a Visiml SearchTask," Journal of Experinmental Psýý,'hology, 51 (1), 1956, pp. 19--24.

68. Green, B.F., W.3. McGill and II. M. Jenkins, "The Time Required toSearch for Numbers on Large Visual Displays," Report No. 36. LincolnLaboratory, Massachusetts Institute of Technology, August 1953, 1.5pp.

69. Ilaeusing, M., ''Color Coding of Ilnfoimation on Electronic Displays,Proceedingslof the Sixth Congress of the Internrational ErgonomicsAssociation, 1976, pp. 210-217.

1I

164 U

BIBLIOGRAPHY (continued)

70. Haines, R. F., "Peripheral and Visual Response Time and RetinalLuminance--Area Relations, " American Journal of Optometry andPhysiological Optics, 52, 1975, pp. 85-95.

71. Ilaines, R.M., L.M. Dawson, T. Galvan and L.M. Reid, "ResponseTime to Colored Stimuli In the Full Visual Field," Report No. NASATN D-7927, NASA Ames Research Center, Moffett Field, CA, March1975, p. 27.

72. Ha-laes, R.F., M.M. Gross, D. Nylen and L.M. Dawson, "PeripheralVisual Response Time to Colored Stimuli Imaged on the HorizontalMedian," Report No. NASA TM X-3086, Ames Research Center,Moffett Field, CA, June 1974.

73. Halsey, R.M., "Identification of Signal Lights: I. Blue, Green, White,and Purple," Journal of the Optical Society of America, 49, 1.959,pp. 167-169.

74. Halsey, R.M., "Identification of Signal Lights: II. Elimination of thePurple Category," Journal of the Optical Society of Amtrica, 49, 1959,pp. 167-169.

75. Halsey, R.M. and A. Chapaxuis, "On the Number of Absolutely Identi-fiable Spectral Hues," Journal of the Optical Society of America, 41,pp. 1057-1058.

76. Ialsey, R.M. and A. Chapanis, "Chromaticity-Confusion Contours Ina Complex Viewing Situation, " Journal of the Optical Society of America,44, 1954, pp. 442-454.

77. Hanes, R.M. and M.V. Rhoad-s, "Color Identification as a Function ofExtended Practice," Journal of the Optical Society of America, 49, 1959,pp. 1060-1064.

78. Harvey, M.E., "An Investigation of Alternative Methods of VisualCueing," Research Report, Texas A & M University, 1971, p. 31.

79. Ileglin, J.,NAVSHIPS Display Illumination Design Guide, Section 11:I-Tuman Factors. Naval Electronics Laboratory Center: San Diego, CA,1973.

165

II1

BIB LIOGHAPHY (continueJ)

I80. Hlennessy, R.T. and G.J. Borden, "Color Perception In the Transitional

Zones of Tricolor Glide-Slope Indicators (GSIs)," Naval Air EngineeringCenter, Philadelphia, PA, AD 771-442, November 1973. I i

81. Hilgendorf, R. L., "Colors for Markers and Signals: lnflight Validation,"AMEL TR-71-77, Wright-Patterson Air Force Base, 011, AD 737-901, j1971.

82. Hilgendorf, A. L. and J. Milenski, "Effects of Color and BrightnessContrast on Target Acquisition, SEEKVAL Project 1A1," AMRI1 TR-74-55, Aerospace Medical Risearch Laboratory, Wright-Patterson AirForce Base, CH, July 1974.

83. Hitt, W.D., "An Evaluation of Five Different Abstract Coding Methods,"numl-an Fctlo....s, 3, 1961, ,•0n "l-110.

84. Jeffrey, T.E. and F.J. Beck, "intelligence Information from TotalOptical Color Imagery," U.S. Army Behavior and Systems ResearchLaboratory, Research Memorandum 72-4, Arlington, VA, Novrember1972.

85. Jones, M. R., "Color Coding," Ifuman Factors, 4, 1962, pp. 355, 365.

86. Jones, L.N., J. Steen and W.JE. Collins, "Predictive \ alidities ofSiev.ra. Clin-ical Color Vis-in. Tests for Aviation Signal Light GunPerformance," Report No. FAA-AM-75-l, Federal Aviation Admini-stration, Washington, DC, AL) 006-792, .January 1975.

87. Kanarick, A.F. and R.C. Petersen, "Redundant Color Coding andKeeping-Track Performance," Iluman Factors, 13, 1971, pp. 183-188.

88. Kelly, K.R., "A Universal Color Lan8uage," Color E12Lineeiing, 3_(2),1965, pp. 1-7. I

89. Klein, T.J., "Night Display--One-Man Aircraft Compatibility Analysis,'Report No. 2-55900-OR-2806, LTV Aerospace Corp., May 1970. I

90. Lee, D.M., "The Conmparative iffectixl'ness of Color In a Sarchi Task,"HIuman Factors Technical Report 69-1, Honeywell Marine Systems Cent vr,Wkest Covina, CA, Januarv 1969.

16b

liiIAk)GlIAl'1iN (continuci)-ý

91. 1 nton, 1%RI, \'[A-V /SEQ I, C1-ew L oading Analysis," ''eport, No.P'AD,7-75209- 401 Crcw Systeins liepartmncal, Naval Air 1)vevlopmevit

Cont er, War-iiinste r, P A, a5'M' 1975.

9 2. 1Lit, A, U., Hi. Young and Mv. Shaffer, 'Sitnle lie;' ci ion 'Iinxet as aFuL-tLý1 On f 1,um in~ane i: for \'arious WaxvelengiIts", ''Percept ionl and

Ujvc~x~scs,10, 1071, pp. 397- :399).

9 3. 1 .oeh, Mi. , .1 , S. Warim and E. A. AIILI isi, Fffets of (0olo', 11iclu1tiVe

lPosition, and tLin' Onset a id O ffset of Signlsl: in; a \'NA cikec"ping Tz'lakh1>;ychmoonmic Sine 9, 1967, pp. 9)5-96.

94. Ydri koff, J1. 1., '''iarge't itkiceognit inn P'tonaneWith Clirom a.tiu anidAcl'romn1at ic. Displays, "I Ion1cywell Syst eflis and ieseare'll Cenlt or,St. Paul, IMN, J1 si inn i 19)72.

115. M1at'sitta, Mi. , et :d., , "'Studies lIn 1Display Symribol I .gilhility 3 Part M1\:'11C I egilhitv of Militiary VIap) Symbllols oin 'ulelevisioti, Repert No.EL:Sl-'lI-66- 315; 'ih MITRE ICor'zporat ion, liedford, IMA. AD 641 -658,Scl~tmni)ti 1 966.

98G. N1 c (-ii, C., N, , I r. and A. C, 1: a re, Coo DI DiteW .'ent al I aurinancv andsuwtn''j'. IM~c,/ hPcumial Ni emorandiun 4-7 1, I luni in Engfinering

I s3.'ot 111it OtS, AbldulPoig( 'cc nt, TOD, April 9,171

97, hRi" anl, Nk! * , "M>igipWa im k >' AW1 SX, t ao r Cu'' "',aLt if'l1ilit'."1Imnat; F~actor's, 7, lDeceninbet 1 965, pp. 52 1-526.

98ý. MICi stem , 1). and 1).,1. Sullix'r,n, ' Guide to I unvin EngneigDsg

fulo isa iDisplAys, "Gonlt ca No. 1"000 14-68- C-0278, EngineeringPscolg 1-ncOfie fNaval Research, onC'

Ai) 693-237, 1968,.

99 Iidto W .IV. 1, lKnowles and G. 1 \V, x's z ci, ''V isal 'l'lMm'es ds

I!tlna th 'ltinal'Puihr 0.w :tid, Green, and Whitle Signal i gli s,"_2AL,1SUu'it'v of 'Lmeict 51, H61, p.54-56.

167

B111,1iOGRAPI IY (continued)

100. Colors, Aeronautical Id Ahts and Lighting Equipmunt, General lieguire -

nients for-, Military Specification MIL -C-25050A (AS(,,, 2 December1 963, with Amecndmneii 1, 30 Septem-ber 1971.. U. S. GovernmentPrinting Office: Washingtoln, D)C.

101. Military Speciflication for Advanced Integrated ModulL'r IInStruwtllell 1at iolSlYsten for a Single Seat Aircraft, Appendix A to A1MIIS1753-101,-15 Sept-'mber 19763,

102. . '1zrew Station Signals, Military Standard MIL-STD-41 1 , 30 Juneu

1W.14 with 1'F-otice 1, 30 August 1974. Department of D)efense: Washington,

103. Morgan, 11.11. and E. A. A11ui,,i, "E ffects of Discriniiniabilil. ,v and

Irrelevant Informat ion on AbSolut 1' .1LudgI1en~t~,-3 PI -Cret ion andC Psy-Cho___

104. Muller, 1',F1., Jr. , h. C. Sidovsk~v, A. J . Slivinske and P'.Mi1. F"itt s,''rhc* Symbolic Coding of Intornmation Oil CaIthode lia'., Tubos and Sili ilarllDisplays, "WAI)C THR 55-375, Wi-igfit Air Development C(,ieite, \Vriglri -

Patl erson Air Force Bas4e, 011. 1!9Th.

1 06. Munns, M. , "'Soint E ott f' T 1i.sp~kv Symbol Varialtion U poll Operat or1Performiaiwe In Aircraft hit e rCt'cetion1," P ercept unald ii( Mot or Skills, 26,1968, pp, 1215-1221.

107. Myer's, V. S., "Aýcounivodzt ioni Eftect'- II \Li ~dti-oloI' I isplayvs, 16'nort

No. AFV1)L TIZ-67- 161, Air' For'ce !'Iigi t 1) vinam -ics L aborat.im-, Ya liPatterson Air Force Base, 011, Al) 826-131, 1)eCtemher 196,Wl

1081. Newman, 1K. * iVand A. 1K. D I Vis, "I7 .(IkiijCl.neitIona No re0W(dun~darit~.n:odn~of \izsun :-SvmIbo1ic I isp tay, i'' , c 11C eport No. 10411, t . S,

NavyE't~isIhrioy C entei, -ý:in 1)ie~gu, CA, Itih' 1:1,61.

109 Odi, 1). J. ., "The 11eriufit:s of Color II mn Aivborwi ntug hi(rated Tacti al

R urir 1;ml, (AP, Alp i-i V177.

11111i*1("GRAPIVY (continued i

110j. Od'dl 1).J ., "The fleulefit s of~ Colur In an Air'borne Integ-al ud >c iaDi~p1:iy SV.3stlern,"Oi'-~y 1elmcl Cor-ferenceX Society fox ullo!.-rl'oalo Disp~lay, S-n D~iego, CA, 23 0ýtolhei' 197~7.

Ill. O pical Society of Au ivrica, Th'le 'Science of Color. Thomuilas Y'. C>vilC4Xnipiwy: TNwv York, N\ , 16953 1563.

11:?. '. stvrimau I-.( ). and Y.C Mulley, ''Advanced Integnilted M)ispbiyS.Sytkem (AT 1)') tot, 'V/SiC.Xrcruft1 '' AIAA / NASA--Amvs V IS!101.Conft reilct, .Tunt, 1977.

113. wiAc'i, Pi c . , ir'* '. olo.' C oding v all~: Indccmlenou Vn \' a )Iv hl

er.~'tn~ '~searc, ' '~cmlog icalI BMW]lei, M, 3136(4, pp. 1 o-M

1 ~4. J'k~-, ., ( C, T1, (remn ;nnd 11,02, B. .ansoih "O~fecI l M I~c1~.'A[.. twi

I~ Po1llnck, .1l J'nc: ni to Pu1Ui .cWent \Yiveleng4is~ mt Vwý-

I ~~iI ~IC~-' IIrceiiti wIIa~d It.~h~ Ma i 25,kw 3, 9f pp. 14 -2?4.

117. Poole, 11.1, judan. ;xans ~ ' m.nbwa sllillgtol!ý 1)C, 1-566'.

1 Il, 'Poston, PN., ''A I .itteral nrt 1evit-w of Cockpit UAgixting," T''echniculMemiorandum 10-7 74,1..S, Amy I lum an I'nginverli g 1laborat ory, Aprill1974.

119. PI-01onuSVI, D).1%1., 'ViSUA1 'Targut Locat ion as a Iuncioa of the ~ NddI~~'

anid Minds of Comln'1 ing SLign 0-s, "~ Jour nal of Lpd ed P:sYcllolokLv, 45 (6G-1961, pp. 420-427.

120. Reed, .1.,'I'lic S3peed and ACCura'Zcy Of IhDinum'lliat inlg Di''l'ClfeiemmS Illthe Hube, B~rilliance, Aiea amnd Shap,1w" Recp()t NO SDC.* --2

Naval Speci-al D)evices (eniter, I orl Washinzwtom, INY, Al) 6391-14%,Sept ellnhOF 1 95 1

i 69

131iILIOGRiAPU1Y (contlined)

121. Rleynolds, 11 N., '''flu X'iOW Effects of Exposure to Electrioltitaiicsceeit

Lighting,"Hna Fatr.* 971, pp 29-40.

1 22. Reynolds, H, R. . M1. White( and R, K. Ililgendorf, 'Detection andRecognition of Colored Sigria 1tighs, " Human Factos, 14, 1972,pp. 227-236.

123, lkizy, E. F.: ''Color Specification for Additive Color Group Displays3,'Report No. HA DC TR-65-278, Pome Air- Developmecnt Center, Griffiss-Air Irc isNY, August) 1965.

124. hvizyo. F.., ''Dihroic Filter S'ecification for Color Additive 1)ispl:tvs;.1!. Further Exploration of Toloerance nkreas and the Influenice of"Other1)isp1a3 Variables," R'eport No. HADC-TR-67-513, USPA? Home Air

- De,,2loprimnt Center, AD 659-346, September 1967.

12V5 COAonh, LI 1 .. tjm ite Cours e of Perceived Sh.-pe a1'1Twilighit I APminlalce'S:

14 Slue, la Greun Spots In Ext rafoveal VisLion,"At Dell,) Iond,-kzionc

it6 (4), 1951, pp. HOW3L4

lot Rusis, Q , 'Ile Utility of Color hin Visual 1)ipls, '" eport No.I±6-1Y/C Sil, Autonetics, Aniaheim, kA, August 19)66.

127. a edn:,w od It uian ý--corsI 16 174, pp. 30-3013.

128. Semple, C. A. , Jr. , H. 3. It apy, Q3. ConWAy, Jr. an1d 1K ,1. liIZhi z'nct ,Anial\'si sf HIunman 1 actors 1)i' a fi Elecct ronic- Fliight I ispiziv Systemns,

Iechinica! Report AZ' AFF' -I) aLi-TO0-! 74, Apil: 1971, 570 pp).

12%. Shont z, XT. ., Q:, . ". t.1 1 l aw I " .\ilin')'A -St oiy of' Visnal~Searzhi Using E; v TI a men t 11 &c.rdings ' M'ur lot'tTdiig fOrz bfo nmhoi 't

Loc-ation," '' Coimac No. PfONHi 4774(001, Office of Naval H~t'sc'urrh,Wasliinigtunl, DC, Decembc'r 1 J"t'.

130. Shonta, IV, 1%, t. A. 1 runmi n~ 1 ., (4 *\ilinis, ' (o1 or Coding fit

Information Location," '' I Iuinui 1 ictors-, 13,v 1 971, pp1. 2-37-246.

IIB LIACGRAPI IY (continued)

131. Shurtletf, D.A., "Design Problems In Visual Displays, Part II: F.aclorsIn the Legibility of Televised Displays, '" Report No. ESD-TR-66-299,The MITRE Corporation, AD 640-571, September 1966.

132. Siegel, M.1I. and A. 1. Siegel, "Color Names as a Function of SurroundL1umi, nance and Stimulus Duration, " Perception and Psyehophysics, 9,1971, pp. 140-144.

133. Smith, S. L. . Color Coding and Visual Search, ' Journal of ExpenimentalPsychology, 64, 1962, pp. 434-140.

134. Smith, S. L., "Legibility of Overprinted Symbols in Multicolored Displays,Journal of Enginecring Psychology, 2, 1963, pp. 82-96.

135. Smith, S. I.-, "Color Coding and Visual Separability In InformationDisplays," Journal ofApplied Psychol og, 4±7, 1963, pp, 358-364.

136. Smith, S, L.. "Visual Displays--' .arge and Small, " Report No. ESD-TDR-62-339, The MITRr_ Corpor lion, for USAF lElectromc SystemsDivision, AD 293-826, 1962.

137. Smith, S.L. and -i. 13. Farquhar, "Color Coding In Fornattc-d Displays,Journal of A id lp s .ology, 4(9, 1965, pp. 393-398.

138. Smith, S.L.. and D. W. Thomas, "Color Versus Shape Coding 1v Infor -mation Displays, '' Journal of Applied.. Pschoi_., 48, 196,t, pp. 137-146.

139, Snadowsky, A. Al., L. F. Rizy and M.F. FElias, "Symbol Identificai ionas a Function of Misregistration in Color Additive Displays, " I't-r"eptualacud Motor Skills, 2•, 1966, pp. 951-"<0.

140. Sperling, 11.G, and C. L. Jolliffe, "Intensity-Time Relationship atThreshold for Spcct-al Stimuli In 1lumnan Vision, J.'Jurnal of thie OpticalSociety of Amwerica, 55, 19d5, pp. 191- 199.

1,1l. Teichner, W.HI., R.E. (Christ and C.Al. Cot-so, 'Color and Co-ding l:, InVisual Displays, " Contract No. N00014-i6-0306, Final Repkcri, Officeof Nrval Research, Washington, I)C, May 1977.

171

13113, I ORAPII Y (continued)

142. Teichner, W.T. and Krebs, M.J., "Laws of Visual Choice ReactionTime," Psychological Review, 81, 1974, pp. 75-98.

143. Tyte, I., J. Wharf and B. Ellis, "Visual Response Times in High

Ambient Illumination," Society of Information Displays Digest, 1975,pp. 98-99.

144. \%ollkx-nann, F.C. and T. Engen, "Three Types of Anchoring Effects InAbsolute Judgment of Hue, " Journal of Experimental Psychology, 61,1961, 11. 7-17.

145. Wagner, D. W., "Color Coding- --An Annotated Bibliography, " ReportNo. NWC TP 5922, Naval Weapons Center, China Lake, CA, March1977.

146. Wagni i v , . LXjtZ.LL tiib XLL ',-0 ,, 'uUig E, L. uI-J.-3,1 3 1'.-j'1

No. NWC TP 5952, Naval Weapons Center, China Lake, CA, JantiuaryI d77.

147, Wagner, ..W., "Targel Detection with Color Versus Black and White

Television," Report No. NWC TP 5731, Naval Weapons Centei'r, China

Lake, CA, April 7.975, 36 pp.

148. Wagnwr, P.W., 'Target Acquisition with Color Versus Black and White"lelevision, " Report No. NWC TP 5800, Namal Weapon-; cntetr, Chinai I9 \d, CA, October 17 "' di

149. Wald, G. , ''lBlue Blindness ln ie Normal F'ovea, Journal of the (_ 'pticalSociety of America, 57, 19-67, pp. 12V-9-1303.

150. \Vnla':cc, W.J., Jr., "!'A lnvestigal ion of T L&g*u l nlmnceinen1 1'sn1 'Colored Backgrounds on a Simn'latcd Radar Display, Mazster' s "'1.iesis,Na-val lPostgradiiate School, Monlterey, CA, AI) 733- 182, Septi-iiit.- 197!o

151. WVeissm-nii, S. and J.S. liInney, ''Rehlaive YeloKW- W.uV Su ,nsitj'ijt y as a

tunctiou of Retinal Position and L~u'ninance icvel, ' ,ournal of thIe

Optical Societ, of America, 55, 1935, pp. 74-77.

172

Sal,

BIBLIOGRAPIIY (concluded)

152. Weitzman, D.O. and J.A.S. Kinney, "Appearance of Color for Small,

Brief, Spectral Stimuli In the Central Fovea, " Journal of tle O2ptcal

Society of America, 57, 1967, pp. 665-670.

153. White, R.M., Jr., M.J. Darnoff and R.I. Reynolds, "Factors Affecting

the Detection and Recognition of Colored Targets, " A1IU TR-72-38,Aerospace Medical Research Laboratory, Wright-Patterson Air Force

Base, OH, May 1972.

154. Whitham, G.I., "The Determination of Display Screen Size and

Resolution Based on Perceptual and Information Limitations, " Infor-

mation Display, July/August 1965, pp. 15-19.

155. Wolf, E. and MVI.J. Zigler, "Some Relationships of Glare and TargetPerception[" WADC-TR-59-394, USAF Wright Air Development Center,

Wright-Patterson Air Force Base. 011, AD 23--279, September 1959.

156. Wong, K. W, and Yacoremelos, "Identification of Cartographic S?'nibolsfrom TV Displays, " Human Factors, 15, 1973, pp. 21-31.

157. Wulfeck, J.W., A. Weisz and \,I. Raben, "Vision In Military kviation,"

Report No. WADC-TR-58-399, Wright-Patterson Air Force Base, Oil,AD 207-780, November 1965.

158. Wyszecki, G. and W.S. Stiles, Color Sc:.cnce. New York: John Wileyand Sons, Inc., 1957.

173 -.

APPENDIX A

DEFINITIONS OF TERMS AND CONCEPTS

YI

AFPENDIX A

DEFINITIONS OF TERMS AND CONCGEFTS

In this appcndix tWe maajor terms Ind concepts used in the Color Display

Design Guide are defined in greater detail. Other selected ternms that are

relevant to ?.iunjn '.s,.on and to displays in general are included. Major

terrmLs appear in alphabctiuAl order. The list below pro-Oides an index to

the contents of mis appendix.

!Ldapraticr. Noise

OCA~IV& zolav process Phosphour

A&m L,,.t Jluinintici Redundancy

A),.;- ilo.--. ?olhr Total redundancy

£r--igti- In eo Parxial redundancy

Resolution

Saturati.onC. 1rrn.a:,i.•i ,. din grainJ__

C r :0inSelf -luminous color

ontz-a-s b'igna).-to-noise ratio

Subt;',ctive color process

. mntr..'.~ -.'ficze -,v Surface colorTarget acquisition

" -k> Detection

. , 3qetlegt-" LocationtI u Identificul ion

1 Jr: irant :zor ''Threshold (visual)

I' . \Visual acuity

"Lu" lt1

I I I I I I I I I I I.5:

V i, bJ a gle

Visual s 2;-itixatv

t'oveal VisionPerilpheral visioln

Wavelength

idaptation

Thte adjustmentC to a new leve) of ambient illumination. When the illumination

is suddenly changed, tIhc eye requires a period of time to become maximally

selliL.ive Jo the ncw luvtl. The greater the change in illumination, the

ionger the adaptaoioni '.me. Adaptation is typically faster to a bright

exivironenwn thbn ,.c- a Jdvrker one. A person going from a normally lighted

room to a li,.-,lar! ti.k root,; may require up to 40 minutes to adapt

completely. '!the reverse -iay r-equire about ten miinutes. In practice

sucin ext.remes se]ý.Ionce- and adaptation to a dark eniviroimi.ent niav

take as little as five to :eri mjnuwe: depending on the difference in lighting.

Anib i ent lilun int li,.a

Light providud ýn thin wor.iti•, a-e- or from an external sou-ce snch as tIIE

sn. Armlbient illuwinat•uC • affi. .•; boll the adaptation of, tl.? & - nd the-

symbhol to background oc..cr:.- or- :fl display. Light -shinintg o)n ýi displ.ay

surtaco will sun with bot ar'i ir 1 gIrround. I'he rcsult will be a lower,

contrrast. High ambient i.lui..i i arn product, display washout 0. e. , the

symnbols can no longer be seen). g

176 r

Additive Color Process

When the additive primaries red, green, and blue are combined, the

radiant energy spectral distribution is the sum of the distributions of the

three primaries. Color television is an example of color produced by

the additive process.

Anomalous Color

A color perception that is markedly different from the physical stimulus.

For example, if yellow is seen at the boundary between a red and green

strip when no yellow is actually present, tile perceived yellow would be

anomalou-. (i. e., out of correspondence with the physical rature of the

area in question).

lirightness

Although this term is often used interchangeably with lightness or lumi-

nance, technically (in visual terms) it more accurately refers to a subjective

impression of relative luminance. That is, a symbol of fixed luminance

will appear "brighter" on a dark backgrowid than on one whose luminance

is very similar to the reference symbol (e. g., a white symbol on a black

vs. a light gray background). Thus, brightness judgments are influenced

by both contrast and absolute luminance.

177

I.

I-.I

Chromaticity IThe color quality of light as defined by chromaticity coordinates of the

Commiss i Internationale de 1' Eclairage (CIE) color coordinate system.

The CIE diagram represents an attempt to specify the various parts of the

color spectrum in terms of three primary colors: red, green, and blue.

These tristimulus specifications were obtained using the data from a large

number of obc.crvers and are expressed in terms of thrcee primary compo-

nents: N (red), Y (green), and Z (blue). Any color in the visible spectrum

can be obtained from an appropriate mixture of these three.

ClhrnnytiCitv .L jaranm

Shown in Figure A.-1, the chromaticity diagram is a system in which all

colors possible with real stimuli are represented within the bounds of the

solid outer line. The axes X and Y are called chromaticity coordinates

and are defined as follows:

x = K/(X + Y + Z)

where X, Y, and Z represent particular proportions of X, Y, and Z,

respectively. Similarly,

y = Y/(f\ YV + Z)

Since X 4 y 4 Z = 1, knowing the value of any two (usually X and Y) deler-

mines the value of the third. Thus, using chrornaticity coordinates, we

have the relative proportions of each of the primaries required to generate

any color.

17178 1

l!

1.0

Y PRIMARY"1931, 2-deq CIE STANDARD OBSERVER

520 50

510 540 -LOCUS OF SPECTRA[ COLORS(WAVELENGTH IN nrn)

0.6

0.6 570Y 5050 PH •~

PHYSICALLY 580

DOSýi6LF noURS YEL'LOW

590

0.4 6004ILLUMINANT 610

C RED 620

400 630 650

700

040k

-I OCUS OF PUiRENONSPECTRAL

47 COLORS X X PRIMARY

460 400

0 0 0.2 0.4 0.6 G.8 1.0

Figure A-1. CIE Chromaticity Diagram

179

~•--.

In Figure A-2, common color names have been assigned to areas within the

diagram. Note that the area represented by different color names is far -

from equal. A practical result of this is that certain colors (e. g., yellow)

must be more rigidly specified and controlled, for display purposes, than

others (e.g., green) to avoid color confusion. It also suggests that adjacent

color names (color areas) should not be used on the same display to avoid T

possible confusion.

The central, partial ellipse around point E in Figure A-2 represents a

"colorless" area, which is seen as shades of white to gray. These data

were obtained using self-luminous light on a dark background.

Color

The characteristic appearance of an object or signal to which the common

labels red, green, blue, etc. are assigned. Color has three dimensions:

hue, saturation, and brightness. (See Figure A-3 and definitions for each

of these terms.) Color is not a property of the object o'r of physical energy

but refers to the perceptual experience of the human observer.

Color Coding

Use of color to convey information on a visual display. Eacl, particular

color represents a value on some information dimension. Color coding

niav be used alone or in conjunction with another dimension such as shape.

Most factors governing the effective application of color to coding apply to jother codes.

I -

lt•U

0.9

0.7 550 --

0.20

0. /GREENqPL

0.5 - 4-A I I _ _ _ _

q0 -58

___________________________________________________ __________ _________ ___qo__

0A _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ 8L__ _ u/ 4 SP C RA RA G0G~ U. 8 . . 3 0.LOR 0. .6 0 .70.

with Colorq Nae7s8ge0taegre

OWS JHT UPIHRED

I IGIlI O(WHiT[, [PA/I ING;)

Li ,

C;L U•

PARK

Figure A-3. Subjective Dimensions of Color

Contrast (Luminance)

A measure of the relationship between the luminance of a symbol and its

immediate surround or backg 'ound. A varlety of dtufferent, mietlho of

expressing luminance contrast are reported in the literature. It is impor-

tant to determine which type of calculation is being used when interpretingI

these results or in evaluatirg ycur own display data. Several of the most

common calculations are presented below.

Contrast (Percent) -- Ratio of the diffe-rence between target and background Iluminance. Peycent contrast may be defined as follows:

a182

0 'b

where

I. oij ect ~L~minance

Ib = background luminance

11- LbI = absolute value of the difference between object and background

luminance

For objects darker than their background, contrast can vary between zero

and 100 percent. For objects brighter than their background, percent

contrast can vary from zero to infinity.

Contrast Efficiency--Another measure of contrast, defined as the ratio of

the sum of the background luminance and symbol luminance and their

differenceL -I,

Max MinL1Va 4 1LMi

where

I Max the higher of the two luminances (symbol or background)

1 the lower of the two luminancesMin

This measure has also been called modulation, contrast sensitivity, and

visibility ratio. IIt can assume values between zero and o'e.

1Meister, D. and D.J. Sullivan, "Guide to Hiuman Engineering Design forVisual Displays," Contract No. N00014-68-C-0278, Engineering Psycholo, yBranch, Office of Naval Hesearch, Washington, DC, AD 693-237, 1969.

1 n'

I

Contrast Ratio--This term has been ust'd ill at least two ways. Ornc is ti•t

simple ratio of one luminance to another: jILuminance ratio = ¼I

We will refer to this as the luminance ratio (1,1R) for the sake of clarity.

A more common expression of contrast ratio is the following:

11 a - I1Mi1

CR = M AlinlMill

Contrast ratio for CI-T displays is a special case where the symbol is

usually brighter than the surround. To cOm!j)pute this contra•St, a frequently

used cquationi that takes ambient liglht into account is

I S 4 W

I,..where

luminance of the symbol measured in anibicnt illuminaiion

C = :C -cn lun-JIi-m•-nu with ýimbiclnt light excluded

Maximum contrast on CRJT displays is typically about 2 to 1 (C 90

percent).

2 Bryden, J. E., 'Design Considerations for Computer-I)riven CRlT Displays, "Computer Desin, March 1969, pp. 38-46° !

IH4 I..,

I.isp.ay

Any device or medium used to convey information to an operator. A visual

display could be as simple as a digital readout or a meter, or as complex

as a battlefield map or sensor imagery onl a CRT.

Dominant Wavelength

A term used to describe the subjective appearance of a color mixture.

Using the chromaticity diagram (see Figure A-4), a line is drawn between

the color expressed in X,Y coordinates and illuminant C. When this line

is extended to the outer ril of the diagram, the wavelength at which it

intersects will indicate its color appearance. The distance along the line

drawn between illuminant C and the outer rim at which the plotted color

appears indicates its purity. Thus, a point three-quarters of the way up3

the line would have a purity of 75 percent.

Hue

The attribute of a color to which commonly used labels such as red, green,

or blue are assigned. The color label assigned to an object or signal

usually corresponds to its dominant wavelength.

3Geldard, F.A., The Human Senses, Second Edition. New York- Wiley,1972.

185

0.9

0.9

520 I

0.6

0.5 980.4 ,600LI!? 61062

600

0.1

410

104___ p I i I I I 0 I I

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Figure A-4. Dominant Wavelength and Purity (Saturation)Defined on Chromaticity Diagram

186

Irre-levanit Color

('oLbi' ad(ed to a dis-play thiat ha~s no task -related mecaning (i. e., coflvcV'S

lio informat ion). If the task- changes, an irrelevant color might become

rclevant. Color thalt is ir-relevant. to a task may have either a neut ral

effect On Operator perfornianc Or it may bt, distracting, producing

pertormniiev d cCrVlenIet.

Luminlance

TheC am11oun1t Of light emitted from a display surface or thic- luminious inten'sity

f ~ l i ''I* L~'l f" .i __ P' ' N I\Inv d i ftf IIt UnIIi t S I hi ve b veen

uISed to qUrIntifyý 111luiallce. The currently acceptable measure is cazidclas

per square wneter (od! in). In Table A-1, conversion factors are pro~ ded

for conmmonily used units. 4 Te subjective measure of luminance is bright -

ness. I'igurt, A -5 provides some common luminance values.

Altunsell Color Notation

Tlhere are several pr'ominent systemis of notation used for- specifying a

particular color. One of the best kniown i-s the Munsell color system. InI

this sYstem, a color notation is obtained by matching a test patch against

a Series of painit Chips. L.Achi col :r chip is specified by three aIlphanumellrlic

4 Bioastironaut ics D)ata 1300k. Washington, D)C: National Aeronautics andSpace Administration, SP-3006.

187

t--C

4c) C. 'CNI C) C) C3

.~4 C -' C)

C' 0 CD 0- 4

',1-' Ck4.

C' 0 0 C) - 0

UPPER LIMIT OF VISUAL TOLERANCE

100000 t

10000 NEW SNOW ON CLrAR DAY OR UPPER SURFACE OF CLOUDS AT NOON

E AVERAGE SKY ON HAZY DAY Al NOON(A AVERAGE SKY ON CLEAR DAYP.- 1000 •wAVERAGE EARTH ON CLEAR DAY

0 - AVERAGE SKY ON CLOUDY DAY cnmw -

10 - AVERAGE EARTH ON CLOUDY DAY 0-0

WHITE PAGE IN GOOD READING LIGHT :>

10 Q6

u WHITE PAPER ONE FOOT FROM ORDINARY CANDLE CL Z

z 0.1S0.1 AVERAGE CHART AT LOWEST READABLE LUMINANCE

._ju Ji SNOW IN FULL MOON 1 1MESOPIC

"o -- - AVERAGE EARTH IN FULL MOON VISION0 1 (RODS AND

c CONES)

00001 - SNOW IN STARLIGHT - -- --- - - - - ---

0.00001 -- GRFFN GRASS IN STARLIGHT

00.000001 -ABSOLUTE THRESHOLD OF VISIBILITY

00

NOTE: 1 FOOT-LAMBERT 1.0764 MILLILAMBERT

Figure A-5. Some Common Luminance Values

189

e rts c'rcsp~~dngto lII(eC lightneCss, 4i1d SMatu rtiOll. ilie ('11 1llt-1n1s

for I lICs diitwiisiolls a re hut" vallue, anld chroxnia, respect ivelY. Figure A -6j

iWthe ve'lt !tts of thle AMUn lli' SYSt eml r11,iot l Tb atZSUSYW

is Used prilmarilly for specifYing sw-ftact Colors.j

Noise

III vlecýtrtoliic eisphwl"S, itoist, Is defiricd ini Iv'i-n oftumiarfled int crft rence

wilth the presentation ofOt, signal produced by graininess or elect rical

inte rferenc e in TV s vstemst. The mnagni u de of thle no0ise relative to Ihle

signa'l al IlnY point" is e~xpre-ssedj for the enilw.xv displayv Ib% thev signal-i 0m-noise

rat1 io.

A less t raditijonal definition of noise in a s vin-oiic or 3AlItanunieric d~pa

kconi~.iers all1 irrelevanlt SY11'abols as dis'play" clutter or nloise. Iffliat is

relevant. (the -;igrial) wvill varY acco rdii ig to the present information require -

menclts of the us er. An ideal displaY Nvot~ld co.-l ai onl~Y the informat ion

required and thus lie ''noise-free.

An inorgyanic wvat erial exhibitfing a1 110ozitL'ie rutaleniSsioi Of elect roni1;1In ICO

rad(iaji~l on uonI el.C: I tionl. 1FhCos'hor-S used t'or S\ýIrelenS Of CH~T-s hae wo

ilimportant cia ract erist ics: color aind persist enee. These characteristics;

Fanrrell, R *j. and .1.M1. Iloothn, ''Deoign I handbook for' ImagerY bit erprentationEqupmet, Boeing Aerospace, CoMpaiwv I-'ecemlber 1971i5.

1I t

~~~d)~(b 10_________

AýiP~~~ U\

F ig u r e ~ ~ ~ ~ VA A - . M n e l C l r S sE m

yo,19i

L

are used to define the phosphor. Phosphor persistence is an important con-

sideration in color selection since, if it is too brief, the dis,!ay may exhibit

flicker. If phosphor persistence is too long, it could interfere with the

subsequent presentation of other colors in the same area. In Figure A-7,

critical flicker fusion frequency (CFF) ci: phosphor refresh rate at which no

flicker is seen is shown for some common phosphors. In that figure the Irelationship between refresh rate and display luminance is shown.

45

35

JO 20 P4

151

W P79

LU

15-

10 P1B909

3.4 34 340

LUMINANCE (cd/rn2

I

Figure A-7. Critical Flicker Fusion Frequency or PhosphorRefresh Rate at Which No Flicker is Seen isShown for Some Common Phosphors

I192

f edundan.

The repetition of !nformation provided in one code by another code. Values

on the two codes are correlated with each other. Redundancy can be total

or partial.

Total Hedundancy--A perfect correlation between values on two or more

codes. Knowing the value on one code provides complete information

about the value of another code. For example, by knowing that all circles

are red and all squares are green, if a given object is a circle, its color

(red) will also be known.

Partial ledundancy--A correlation between values on two coding dimen-

sions in which only limited overlap occurs. Knowing the value on one

code does not completely determine the value on another. Fol- example,

a digital readout may be color-coded as red a high, greeii - medium, and

yellow ý low. Knowing the color of the r( idout will specify a range of

niumbers but not the exact numerical value.

Resolution

In a CRT-typc display, resolution is defined in terms of the number of

line-space p irs that can be seen per unit linear dimension. In Figure1

A-8, a standard resolution target is shown. Patterns of this general

type are used to measure the resolving capability of optical systems.

Television systenms are generally designed to have equal horizontal and

vertical resolution. Vertical resolution for an interlaced TV system

equals the number of active lines per frame times 0. 7, In normal room

193

I

-2 -I

2 =E Iii .. ,• 'III- 1111=2

U = I I I i' Iu: -

3SAF 1951 E

E

FigCure A-8. Standard Ilesolution Target

light the average eye can discriminate 40 pirallel lines alternating black 1

and white (i. e., 80 TV lines) per degrec of arc.

Saturation I

The relative purity of a color defined in te'ms of its departure from a

white or gray of the same lightness. Fo' example, the two colors labeled1

pink and red have approximately the same hue but the red would be highly

saturated and the pink would be low in saturation or desaturatcd. Zero

saturation colors are black, gruy Land white. (See ligures A-3 and A-4.

19

S~I

194

I'.

Solf-I, urrinous Color

Typified by a color CR'T, the color sensation is not produced by reflected,

ambient illumination; the symbol itself is the light source. On a CRT,

different colors are produced by using different phosphors. Because of its

lack of dependence on reflected light, a self-luminous color display can

be viewed over a wider range of ambient illumination than can a surface

color display. At high levels of illumination, however, colors may not

be visible because of reduced contrast. Color of ambient illumination

has less effect on self-luminous displays t!,an on surface colors.

Signal-To -Nois c Rlatio

Typically defined as tl'e relationship between signal magnitude and noise.

Noise is usually thought of as anything that is displayed that is net part

of the signal and thus contains no information. In projected displays

such as TV, graininess or electrical interference is present in every

frame. The signal-to-noise ratio (SNR) for such systems can be quantified

and compared to desired values. An SNR of :iGd. is considered goad

quality fox' television systems. The SNH required depends on the object to

be detected: a bar-type resolution pattern requires an SNP% of 3:1 to be

visible.

Subtractive Color Process

Color achieved by mixing dyes or pigments that selectively absoib the

radiant energy in a portion of the visible spectrum. Color photographs

and various other surface colors are achieved by this process.

195

I

Surface Colors iAs the term implies, surface colors are an integral part of the object

(examples include photographs, printed map3, decals, paint). Color is

determined by light reflected off the surface of that objeo,. Any surface

absorbs certain wavelengths of the light shining on it und reflects others. !

The one(s) reflected determine(s) the pt rceived color. For example, if

white light is directed at an object and the object absorbs all but the longe.st 3wavelengths that it reflects, the object will be seen as red. The appearance of

a surface color may vary drastically if the color of the ambient illumiina-

tion is changed. Surface colors are seen only at moderate to high ambient

illumination in the photopic range (see Figure A-5',i. low this range, Iobjects appear colorless or as shades of gray. ITart Acguisition

A general term us d to tescribe the process of searching for ýad/or

labeling a target, The following subtasks may or may inot all be part of

this process depending on the situation: 1Detection--Determining that a given sigtml or target is present when its

onset or occucren'e is either not expected or is uncertain. 5Loeation--Determinir:g the position of a target in atli iiforimatted display I0or among randomly positioned non-target objects.

S1I

Identificotion--Labeling a symbol or shape by a name that has task-related

meaning or determining the status of some variable by reading arid inter-

preting a coded message.

Threshold (Visuan.l)

A measure of maximum sensitii,jty to a stimulus. For example, the lowest

luminance at which a small spot of light can be seen is the luminance

threshold. Usually a threshold value is reported at the point where the

stimulus is seen 50 percent of all the times it is presented, unless another

percentage value is specified. In adapting such data to operational systems,

care 3hoald be taken to revise the threshold value upward to permit 99 to

100 percent detectability under typical operational viewing conditioni. If -

the standard deviation (SD) is given, the 50 percent threshold value should

be increased by at least three SDs.

Visua 1Acuit_

The ... inimum. det.ail rcso..1vb1c by the human eye. Thi-s bhsolute thre.shold

will vary depending upon such factors as signal luminance, contrast,

signal duration, anJ, to a more limited extent, signal hue or wavelength.

Acuity is measured in a variety of ways. One common device is the

Landolt Ring shown in Figure A-9. The Landelt Ring has a break in it

at one of four possible lo--ations. The observer indicates where the

break is (top, bottom, left, or right). The smallest opening that can be

detected in the ring is an indication of visual acuity. Another measure of

acuity, the Snellen Eye Chart, is similar to the Landolt Ring, except the

observer must identify letters viewed at a fixed distance.

197

~I

jWE

I :,N I I I I N I

t~itN AI, ANP• ý'.ll I I hAlti

I AND)OL I Ci

jW IS 'IMH TARGI I STROKM WfIOTIH'

Figure A-9. Common Visual Acuity Test Targets

1981

I

Another type of acuity measure is the minimum separation acuity defi,,ed

as the smallest separation that can be resolved by the eye or by a given

display medium.

Visual Angl

The size of the object at the eye of the observer. Visual angle takes into

account the actual size of the object and the distance of that object from

the eye. It is usually expressed in degrees, minutes, or seconds of arc.

Equation (A-1) below can be used to calculate the visual angle subtended

by an object or symbol of known size at a specified distance. Using

Equatio. (A-2), the symbol size required to achieve a specified visual

angle can be calculated.

Visual angle in degrees = 2 arctan h/2d (A-i)

where h = linear symbol dimension (height)

d = distance from eye measured perpendicular to line of sight

A close approximation to Equation (A-i) can be computed for angles of6

less than ten degrees (600 minutes), using constants as follows:

Visual angle in degrees = d 3hd

where h and d defined as in Equation (A-i)

6 Baker, C. A. and W. F. Grether, "Visual Presentation of Information,"Wright Air Development Center, WADG-TR 54160, AD43-064, 1954.

199

To convert from angular to linear measure

h = 2d tan visual angle/2 (A-2)

in Table A-2, some typical values are provided as examples of the relation-

ship betwen size, viewing distance and visual angle.

TABLE A-2. SYMBOL SIZES REQUIRED TO ACHIEVE GIVEN VISUALANGLES AT SEVERAL VIEWING DISTANCES (Cell values,in parentheses, represent symbol size in inches and centi-meters.)

14 V ag I lisof ;tro

___II _iIIIet5 10 i5 1 21 30 40 • 45

b 01 0.05 0.10 0.16 0.22 0. M1 0.42 0.47(0. 13) (0, 27) (0. 40) (0. 56) (0. 810) (1. 06) (1.20)

3 81 0. 0,5 0. 09 .1 1 ..20 0. 28 0. 37 o. 1,2( .12) (0). -24) (0. : 5) (0., 50,) (0. 71) (o). (I5 ( . o(;)

2 8 71 o.0. O . 0li 0. 12 0.17 0. 24 0,3 0. 0.370( . 10) (0. 21) (0. :1) (0. 43) (0. 62) jo. i' , I .,,,

61 j . 03 9.07 o. 10 0. 15 0. 21 0. :M8 0. 31(0. 09) (0. 18'!) (1). 27) (0). 35) (0. 53) (0. 71) 0. 'il)

28 51 0 0 o. o6 0, 0•) 0. 12 0. 17 0.2" 0. 2i(0. 07) (0. 15) (0. 22) (0. 1) (0. 44) (0. 5!') (0. 68)

2I

2003

I

Visual Sensitivity

Tihe human eye is divided into two ni,&jr areas. fovea arid periphery, each

of which has different levels and types of sensitivity to visual stimulation

(see Figures A-10 and A-11.)

NASAt SIM I I t MPORAI SIM)t

•-COH•NIA

| s. R ET INA-- PtH. L')OMI N' A I E

,P W, NE FIVE-

CON LS PH t DOMI N,,AT 7L

Figure A- 10. Horizontal Cross Section of the Right Eye

201

I-O'I

I.... -.. .. ,f~- • I IN ,,'l S cN1L --

S-- 0 A I IIV I VISUAklSIA I "I I 1

U E T .".o ->-

D Z I I ROM I MI

.. . . - It "- , ,

a~o A

$I)0 --- - , + -0r f/ "I , I I

Sc f I ,

m-xnUlinthis Th0e fovea cotins onlyl cones. l)

flet GFitrt tI Ira dM o lth F ('V- A

Figure A-11. Reiationship of Visual Acuity to the Ifistributionof Rods and Cowes

Fovea]. Vision- -Foveal vis ion cove'rs ,n area within :about + 1 degree from

the center cf the line ot sight; both visuatl acuity anti tolor v'isionl are'

IwLxinlum in• this at'ea. The fovea contain~s only cones.

Per-ip'heral Vision- -The remallindter of the visua.l field that lies oatside the

fovea]. field of view. Sensitivity to lovy levels of light is m~uc'h greatcr in

the periphery. Acuity i~s substantially lower , however, so, hle periphery

is not sensitive to fine detail, Color sensitivity also decreases as the

signal moves out into the purip'iery. The peripherY contains both rots and

cones but the number of cones decr"aSCS MIar'kcclly aIs the diStaInC t'rom-

the fovea increases. Peripieral vision extends 180 degrees fromn the line

of sight anti about 130 degrees in the vCrtical direction. (see Figurls A-12 jand A-13. 1

I202 I

e K

1200E)G 'rELLOW 6

i0t DEG BLUE .

60LEC, GREENRID R

HOM4IZONIALLY

Figure A-12. Hlori?0ontal Angulklr C'olor Lir.mits

NWC I P 5 922

130 A) M.G V041l t

OI 2 G YE LLOW

V f•TICALL.'t 4% DEG REDO

40 DEG C.;E'tN

Figure A-13. Vertical Angular Color Limits

203

These figures also indicate that the visual field is maximunm for white

signals, and that the field for color signals is more restricted, with

green and red having much smaller areas of sensitivity than other colors. IWavelength 1A specific point on the visible portion of the spectrum (see Figure A-14).

Usually used to describe the characteristic of a signal of some specified Jhue as shown in Figure A-14. Most real-world signals are not produced

by one wavelength but are a mixture of more than one. The particular

color seen in such a mixture is determined by the dominant wavelength.

b

ii

Iii

II

Ii204

I 'I

LjJ

C0 CDCC 0 l CD 0

I.-4r4- - -Y

M -. ")Le~ ~ .- 00 )2: C-4 '-D C

LUI LUJ LU

rc t <tC

- 04

V/)

c~, jul I I b

I II

LU C) LUDLw) .-.J C: C0-UC L L e

>-J 2=

-JL

of) -J0

A: C)(A ( CC

C))V) LU

LUC L-UJ j LL

C-.)-. I LUI

LLLU

F--

205

d

APPENI)L\ IREFEE'NCES

±. Meister. D. and D. J. Sullivan. "Guide to llunIan Engineering Designfor Visual Displays, " Contract No. N00014-68-C-0278, EngineeringPsycholog5 Branch, Office of Naval Rcsearch, Washington, D)C.AD 693-237, 1969.

I2. Bvyden, J. E., "Design Co.iiderations for Computer-Driven CRT

Displys, " Coinputer__Desi•n, March 1969, pp. 383-46.

3. Geldard, F. A. , The Human Senses, Second Edition. New York: Wile.,19t2.

11. Bioa:stronautics Data Book. Washington, DC: National AeronauticsanJ Sp ce Auministration, SP-300G.

5. Farrell, 11. J. and J. Al. Booth, "Design t1andwook for imagery Inter--pretation Equipment, " Boeing A,,rosF'tce Company, December 1975.

6. Bakei, C. A. and W. F. Grether, "Visual Presentation of Information,Wright Air Development Cent2,r, WADG-TR-54160, AI)43-064, 1954.

LIII!|I.

206 .

DISTRIBUTION LIST

Chief of Naval Research Office of the Deputy Under800 North Quincy Street Secretary of DefenseArlirtton, VA 22217 OUSDRE (E&LS)Attn: Codes 221 7 Pentagon, Room 3D129

455 1 Washington, DC 20301441 1 Attn: CDR ChatelierlOOM 1H

HeadquartersDefense Documentation Center Department of the NavyCameron Station Naval Material CommandAlexandria, VA 22314 12 Washington, DC 20360

Attn: Systems EffectivenessDirector Branch MAT 08T21 1Naval Research Laboratory 98T24 IWashington, DC 20390Attn: Code 2627 Commander

Naval Air Systems CommandCommanding Officer Washington, DC 20360Office of Naval Research Attn: AIR 5335 1

Branch Office 5313 1New York Ar-ea 340715 Broadway (5th floor) 340F INew York, NY 10003 360A 1

03PA 1Commanding OfficerOffice of Naval Research Coqmander

Branch Office Naval Sea Systems Command1030 East Green Street Washington, DC 20360Pasadena, CA 91106 Attn: NSEA 0341

Commanding Officer CommanderOffice of Naval Research Naval Electronic Systems Cofmand

Branch Office WashinyLon, DC 20360Building 114, Section D Attn: EL[X 304 1666 Summer Street 4701 1Boston, MA 02210

Naval Facilities EngineeringCommanding Officer CommandOffice of Naval Research R&D Plans and Programs

Branch Office Code 03T536 Clark Street Hoffman Building IIChicago, IL 60605 1 Alexandria, VA 22332

Attn: Mr. M. EssoglouOffice of the Chief of Naval

Operations Commanding OfficerDepartment of the Navy U.S. Naval Air Development CenterWashington, DC 20350 Wanrinster, PA 13974Attn! OP-986D 1 Attn: Codes 20P4 1

OP-987 1 604 1OP-506 1 607 1

5023 1Tech Library 1

I

Corriandiy OficerComm~anderCommanding Officer Naval Ship Research & Development

Human Factors Section Center, Annapolis Division

System-i Engineering Trst H);man Factors Engineering Branch

DirectorateU.S. Naval Air Test Center

Annapk is, MD 21402

Patuxent River, MD 20670 CommanderNaval Coas~al Systems Laboratory

Bureau of Mdedicine & Surgery Code 712

Aerospace Psychology Branch e City, 1L 32401Code 513

Pnm iy * 20

Attn. CoR R. Gibs-01 Naval Training Equipment Center

WAs lttn: CRR 2037- Orlando, FL 32813

Naval Medical R&D Command Attn: Technical Library

Code 44 N-71

Naval Medical CenterBethesdca, Cente 20Navy Personnel Research &

Bethesda, MD 20014 Development Center

Attn: LCDR Robert giersnler Code 305San Diego, CA 92r52

Submarine Medica) Research Lab Attn: Code 311

Naval Submarine Base

Groton, CT 05340 M Human Factors Engineering Branch

ottn: Dr. rJ. A. S Kinney Pacific Missile Test Center

3 .aPoint Mugu, CA 93,42

terospace Psychology Division Attn: Code 1226

Naval Aerospace Medical InstitutePensacola, FL 32512 1Dean of Research Administoatiofl

Naval Postgraduate School

Comanader Monterey, CA 93940

Naval Ocean Systems Center Oean of the Academic Departments

San Diego, CA 92152 U.S. Naval AcademyAttn: Code 7113

U., aa cdm

8231 1 Anni.polis, MD 21402

Corniander Operations Research . .Department

Naval Weapons Center Naval Postgraduate School i

China Lake, CA 93555 Monterey, CA 93940

Attn: Code 3175 3 Attn: Dr. Gary Poock

CoanderanderNaval Surface Weapons Center

Naval Underwater rS ystes Center

Dahlgren Laboratory Department SB 324

DahlrenVA 2910Newport, RI 02840

Dahlgren, VA 20910

Attn: lechnicel Library 1 Display Branch

Conmmander Naval Underwater Systems Center

Naval Avionics Facility New Londcn, CT 06320

6000 E. 21st Street Attn: Code TD112

Indianapolis, IV 46218

Attn: Technical Library i

U

]IDirector, U.S. Army Research Director

Institute Human Engineering Labs5001 Eisenhower Avenue Aberdeen Proving Grounds, MD 21005Alexandria, VA 22333 1 Attn: AMXRD-HEL

U.S. Army Avionics Research & Air Force Avionics LaboratoryDevelopment Activity Air Force Systems ConTand

Attn: DAVAA-E Wright-Patterson AFB, OH 45433Fort Monmouth, NJ 07703 1 Attn:. AFAL/RWI

U.S. Army Electronics Research Aeronautical Systems Division& Development Conmmand Air Force Systems Command

Attn: DELET-BD Wright-Patterson AFB, OH 45433Fort Monmouth, NJ 07703 Attn: ASD/RW

ASD/AERSCormandant, U.S. Marine CorpsHeadquarters, U.S. Marine Corps Aerospace Medical ResearchWashington, DC 20591 LaboratoryAttn: RD-i Wright-Patterson AFB, OH 45433

Attn: AMRL/HEAChief, C3 DivisionDevelopment Center Air University LibraryMCDEC Maxwell Air Force Base, AL 36112Quantico, VA 22134

Dr. Gordon EckstrandCommandant AFHRL/ASMU.S. Coast Guard Headquarters Wright-Patterson AFB, OH 45433400 7th Street, NWWashington, DC 20591 Air Force Office of ScientificAttn: GDST/62 TRPT Research

Life Sciences DirectorateCommanding General Bolling Air Force BaseU.S. Army Material Command Washington, DC 20332Washington, DC 20315Attn: AMCRD-HA Headquarters, Rome Air

Development CenterUHQ, D+•en.t f te Ar-yAir F •,,-,m C^nnDAPE-PBR Griffiss Air Force Base, NY 13441Washington, DC 20546 Attn: RBRACAttn: Mr. J. Barber

Federal Aviation AgencyDirector, Organizations & NAFEC Bldg. 10

Systems Research Laboratory Atlantic City, NJ 08405U.S. Army R2search Institute Attn: Code ANA-2305001 Eisenhower AvenueAlexandria, VA 22333 Defense Advanced Research Projects

AgencyU.S. Army Aeromedical Resedrch 1400 Wilson Boulevard

Laboratory Arlington, VA 22209Attn: CAPT Gerald P. KruegerFort Rucker, AL 36362 1

Institute for, Defense Analysis IBM Watson Research Center400 Army-Navy Drive P.O. Box 218Arlington, VA 22204 Yorktown Heights, NY 10598Attn: L. Biberman Attn: Ifay Chang

Office of Life Sciences North Hills ElectronicsNational Aeronautics & Space Alexander Place

Administration Glen Cove, NY 11542600 Independence Avenue Attn: S. SherrWashington, DC 20546Attn: Dr. Stanley Deutsch 1 Xerox Corporation

Palo Alto Research CenterNational Oceanic & Atmospheric 3333 Coyote Hill Road

Administration Palo Alto, CA 9430411400 Rockville Pike Attn: B. KazanRockville, MD 20852Attn: Dr. J. Miller Dr. Marjorie Krebs

625 Euclid AvenueUniversity of Illinois Erie, PA 16511 5Coordinated Sciences LaboratoryUrbana, IL 61801 Tektronix, Inc.Attn: Dr. G. Slottow P. 0. Box 500

Beaverton, OR 97005Virginia Polytechnic Institute Attn: A. SilzarsUn jJ. vZ ,LI bV l caiyrir e 2r-1

Blacksburg, VA 24061 K. ConsidineAttn: Dr. H. L. Snyder

Sperry Flight SystemsHoneywell, Inc. 2111 N. 19th AvenueSystems and Research Division M/S 109-C2600 Ridgway Parkway Phoenix, AZ 85302Minneapolis, MN 55413 Attn: J. R. Trininier 1Attn: Dr. A. Kanarick 1

Jim Wolf 1 Northrop Electronics Division2301 W. 120th Street

General Electrir Hwhne, CA 9n02n

Research and Development Attn: Walt GoedeBox 43Schenectedy, NY 12301 Westinghouse Electric Corp.Attn: J. E. Bigelow Research and Development Center

Pittsburgh, PA 15235Magnavox Company Attn: Dr. Peter BrodyAdvanced Technology GroupFort Wayne, IN 46804 Xerox CorporationAttn: Dr. C. Craighead 1 Webster Research Center

Paul Halberg 1 800 Phillips Road W114Webster, NY 14580

Kaiser Aerospace and Electronics Attn: J. B. FlanneryCorporation

1651 Page Mill Road Lucitron, Inc.P.O. Box 11275 Sta, A 1918 Raymond DrivePalo Alto, CA 94306 Northbrook, IL 60062Attn: G. Carroll Attn: Alan Sobel

I.

Boeing Commercial Airplane Co. Richard E. ChristP.O. Box 3707 Box 3UJ - New Mexico State UniversityM/S 47-09 las Cruces, NM 88003Seattle, WA 98124Attn: A. F. Norwood 1 Jan Wirstad

Ergonomrad ABPHarris Corporation Box 100 32Electronic Systems Division S-650 10 KarlstadMSI/1821 SwedenP. 0. Box 37Melbourne, FL 32901 Professor Dr. R. BernotatAttn: Terry Riley 'Forschungsinstitut fur Antropotechnik

BuchstrasseLockheed Aircraft Corporation 0-5309 MechenheimP.O. Box 551 Dept. 96-26 West Germany FRGBurbank, CA 91520Attn: Don Oda Albert G. Bowman

Human Engineering Grouprelco Electronics Institute fur Perception TNODepartment 3210 Kampweg 5, P. 0. Box 237929 South Howell Avenue Soesterberg, The NetherlandsOak Creek, WI 53129Attn: Earl Strandt Professor Dr, C. Graf Hoyos

Lehrstuhl fur PsychologieHewlett Packard Technical University of Munchen1000 NE Circle Boulevard Lothstrasse 17Corvallis, OR 97330 8000 Muchen 2Attn: Paul Van Loan West Germany, FRG

Department of Psychology Chris PoultonThe Johns Hopkins University MRC Applied Psychology UnitCharles and 34th Streets 15 Chaucer RoadBaltimore, MD 21218 Cambridge CB2 2EFAttn: Dr. Alphonse Chapanis 1 England, U.K.

Department of Engineering Administration Ronald S, EasterbyGeorge Washington University Applied Psychology DepartmentSuite 805 University of Aston2101 L Street, N.W. Gosta GreenWWashington, DC 20037 Birmingham 84 7ETAttn: Dr. Meredith P. Crawford 1 England, U.K.

Oceanautics, Inc.Depart'oent of Ocean and 422 6th Street

Electrical Engineering Annapolis, MD 21403Massachusetts Institute of Attn: Dr. W. S. Vaughan

TechnologyCan;ýridge, MA 02139 Naval Ocean Systems CenterAttn: Dr. Arthur B. Baggeroer 1 Hawaii Laboratory

P.O. Box 997Kailua, Hawaii 96734Attn: Dr. Ross L. Pepper

Department of Psychology Panel Displays IncorporatedVanderbilt University 211 South Hindry AvenueNashville, TN 37240 Inglewood, CA 90301Attn: Dr. Robert Fox 1

RCA LaboratoriesHuman Factors Research, Inc. David Sarnoff Research CenterSanta Barbara Research Park Princeton, NJ 085406780 Cortona Drive Attn: Phil HeymanGoleta, CA 93017 I iAttn: Dr. Robert R. Mackie 1 Perceptronics, Inc.

6271 Vatiel AvenueDep rtment of Mechanical Engineering Woodland Hills, CA 91367 jMassachusttts Institute of Technology Dr. Cershon WeltmanCambridge, MA 02139Attn: Dr. T. B. Sheridan 1

Department of PsychologyGartley HallUniversity of Hawaii at ManoaHonolulu, Hawaii 96822Attn: Dr. William R. Uttal 1

Control Data CorporationResearch and Advanced Design Lab.4290 Fernwood StreetArden Hills, MN 55112Attn: U. B. Bonstrom I

Boeing Aerospace CompanyResearch and Enqineering DivisionP.O. Box 3999 MS 41-08Seattle, WA 98124Attn: W. J. Hebenstreit 1

Hughes Aircraft CompanyDisplay Systems & Human Fautors

DepartmentAerospace GroupCulver City, CA 90230Attn: W. Carel 1

McDonnell Douglas CorporationP.O. Box 516St. Louis, MO 63166Attn: H. F. Engineering 1

R. W. Fisher 1G. Mills 1G. Adam 1

32

Ut.